A historical survey up to Chomsky and Tesnière
Prepublication draft
February 22, 2025
[1.1] At least since the late 1950s, with the publication of Chomsky (1957) and Tesnière (1959), the idea to draw graphical representations of sentence structure has become a cornerstone of contemporary linguistics. However, graphical procedures for the syntactic analysis of language have a long history, far predating Chomsky and Tesnière. Both these men have been quite tight-lipped about the sources of inspiration for their syntactic displays, implicitly suggesting that their graphical conceptualisations of sentence structure were singular innovations. Quite to the contrary, in this book I will attempt to show that there is a continuous tradition of graphical grammar going back to about 1830, with even a few incidental precursors predating that starting point. This history, about 200 years at the time of this writing, has many twist and turns, with forgotten, rediscovered and independently discovered ideas, with periods of frantic innovation followed by phases of settling tradition, and in general with the promise of graphical grammar lingering in the scientific air.
[1.2] A central driver of innovation for visual methods throughout this history was education. Time and again, the didactic desire to clarify the often twisted linguistic constructions used by our most beloved geniuses of language, by our great authors and poets, drove teachers to the drawing board. Quite literary, it might very well have been the general availability of blackboards in schools that kicked of the graphing craze starting in 1830s. Lyman describes for the USA that “as late as 1835 the idea of using slates and blackboards was exceedingly novel; in only a few schools does it appear to have been attempted before this time” (Lyman 1922: 148). He even quotes a contemporary educator from the 1830s:
In the study of grammar the blackboard may be used to exhibit the inflections of the various parts of speech; it may also be used in syntax, to point out the connection of the principal words to each other. The method of doing this is by writing on the board the sentence to be parsed, and then connecting by curved lines those words that have any grammatical connection with each other. The instructor at the same time pointing out what that relation is. (James Ray 1830, cited in Lyman 1922: 148, emphasis added)
[1.3] Yet, also from the very start the graphical methods were also used to clarify linguistic structures in scholarly debates.
[1.4] Previous studies into the history of graphical grammar: (Gelbe 1880; Arens 1902; Grosse 1966; Brittain 1973; Percival 1976; Stewart 1976; Baum 1976: 36-42; Coseriu 1980; Forsgren 1992; Thümmel 1993; Seuren 1998: 219-227; Franzkowiak 2013; Seuren 2015; Dukes 2024)
[1.5] Studies on the history of grammar writing related to schools in general: (Lyman 1922; Glinz 1947; Huston 1954; Leitner 1991; Görlach 1997; Knobloch 2000; Graffi 2001; Linn 2006)
[1.6] The current structure of this manuscript has all historical developments organised roughly by nationality. At this stage, this is just an practical strategy to classify the material. The ultimate structure of the book might very well be organised by scholarly traditions. However, the basic units of the story will always be the third-level sections that summarise the contribution of a specific author. The general plan is to compile for each author a little miniature of a few pages with a few selected images illustrating their graphical approach. Only in very few cases I have decided to discuss the same author in multiple sections.
[1.7] There are often multiple editions of the works that are discussed here, and these different editions sometimes contain quite different ideas. So it is important to check exactly which edition is consulted. The bibliography includes links to scans of the originals, with different entries for the different editions consulted. Unfortunately, not all crucial editions are available online yet. For now, low-quality scans of crucial pages from books that I have been able to inspect in person are linked via a personal repository.
[1.8] As a result, every cited reference in the text is clickable and leads directly to the bibliographical entry, which contains a link to the original online. This will be exactly the edition that is cited in this book, so the cited page numbers will match. You will still have to manually scroll to the respective page numbers, but in this way the original context of every contribution discussed here is easily available for closer inspection to the interested reader.
[2.1] Slowly, during the 17th and 18th Century, a new terminological amalgamation arose, combining terms from Aristotelian logic with concepts from the descriptive grammatical tradition. The result of this mixture are the contemporary terms “subject”, “object” and “predicate”, which seem to be first used in the modern sense at the start of the 19th Century. The history of the emergence of these terms deserves a detailed investigation in it’s own right (cf. Forsgren 1985; 1992 for the German grammatical tradition). The following summary is but a very short and strongly simplified overview of this terminological development. The underlying impetus for this development appears to be an ambition on behalve of grammarians to insert more “logic” into their analyses, or, in more modern terms, to create general terms for grammatical functions on top of the description of the grammatical form of linguistic elements.
[2.2] Everything starts with Aristoteles. First, the term “subject”, through Latin subiectum, is a calque from Greek ὑποκείμενον ‘that which lies beneath’. This term is one of the central tenets of Aristotelian philosophy of substance. Similarly, the term “predicate” is derived from Latin praedicatum, which is a calque from Greek κατηγορούμενον ‘that which is being spoken against’. These two elements, subject and predicate, are the building blocks of the Aristotelian (and medieval) logical proposition, not of a linguistic sentence. In contrast, linguistically a Greek λόγος ‘sentence’ consists of ὄνομᾰ ‘noun’ and ῥῆμα ‘verb’. Although Aristotle describes a third element linking the logical terms ὑποκείμενον ‘subject’ and κατηγορούμενον ‘predicate’, he does not use a specific term for this third element. Moro (2017: 35) argues that only with Abelard in the 12th Century a specific term is introduced for this link, namely the Latin term copula. Separataly, the term object, from Latin objectum, is a calque from Aristotle’s ἀντικείμενον ‘that which lies opposite’. This term does not play any role in classical logical analysis, nor in linguistic analyses. It only figures as a philosophical concept for observed reality.
[2.3] In the ensuing centuries, the Latin terms subiectum, praedicatum and copula become a staple in medieval logic, specifically in the analysis of the syllogism. In contrast, these terms are never used for grammatical analysis of language. For example, the 13th Century scholastic grammarians (the “Modists”) use suppositum and appositum approximately in the sense of the Greek ὄνομᾰ and ῥῆμα, respectively (Robins 1980: 234).
[2.4] In the 17th Century, the Grammaire Générale et Raisonnée from 1660 by Antoine Arnauld and Claude Lancelot is probably one of the first occurences of new terminology. In their attempt to link language to logic and thought, there is a first glimpse of a terminological confluence of Aristotelian logic and the grammatical tradition, as the french word sujet ist used, although alongside attribut and not yet something like prédicat:
La connoissance de la nature du Verbe dépend de ce que nous avons dit au commencement de ce discours, que le jugement que nous faisons des choses (comme quand je dis, la terre est ronde) enferme necessairement deux termes; l’un appellé sujet, qui est ce dont on affirme, comme, terre; & l’autre appellé attribut, qui est ce qu’on affirme, comme ronde: Et de plus la liaison entre ces deux termes, qui est proprement l’action de notre esprit qui affirme l’attribut du sujet. (Arnauld & Lancelot 1660: 89)
(‘The knowledge of the nature of the Verb depends on what we said at the beginning of this discourse, that the judgment we make about things (as when I say, the earth is round) necessarily includes two terms: one called subject, which is what we affirm about, such as earth; and the other called attribute, which is what is affirmed, such as round. And moreover the connection between these two terms, which is strictly the action of our mind that affirms the attribute about the subject.’ translation from Kahane 2020: 100)
[2.5] Also in the 17th Century, Christopher Cooper in his Grammatica Linguae Aglicanae, a grammar of the English language written in Latin, makes an attempt to bridge the two worlds of grammatical and logical analysis, clearly articulating the difference between the grammatical term “nominative” and the logical term “subject”. He also uses the term “predicate”, but apparently not yet “object”:
Substantivum … Et hoc a grammaticis noncupatur casus nominativus, a logicis subjectum. … Id … vocatur praedicatum; quod si duobus verbis exprimitur, substantivum vel adjectivum subsequens copulam in regimine. (Cooper 1685: 171-172)
(‘Noun, … grammatically it is called nominative, logically it is called subject. … It is called predicate when there are two words expressed, noun or adjective following a copula.’)
[2.6] In the 18th Century the use of the logical terminology for functional categories in the analysis of sentence structure broadens. For example, in France, l’Abbé Girard in Les vrais principes de la language françoise (1747: 90-92) uses the terms Subjectif, Attributif and Objectif (see also Forsgren 1985: 43). He seems to build on the sujet/attribut terminology from the Grammaire Générale et Raisonnée (see above) and extends this functional approach to other aspects of linguistic structure, notebly introducing the term “object”. In Germany, Johann Christoph Adelung in his Umständliches Lehrgebäude der Deutschen Sprache (1782: 61) uses the terms Subject and Praedicat, but not anyting resembling the term object. In yet another example, this time from England, Joseph Priestly in his Rudiments of English Grammar (1772) uses the terms “subject” and “object”, but he does not appear to have used the term “predicate”:
Q. What is meant by the Subject of an affirmation?
A. The person of thing concerning which the affirmation is made.Q. What is a verb transitive?
A. A verb transitive, besides having a subject, implies, likewise, an object of the affirmation, upon which its meaning may, as it were, pass; and without which the sense would not be complete. (Priestley 1772: 13)
[2.7] In the 19th Century the use of the terms subject, object and predicate becomes standard practice. For example, Heyse (1827: 633-634), a highly influential German grammar from the start of the century, uses the terms Subject/Gegenstand, Prädicat/Äußerung and Object/Zielwort. However, even he feels the urge to explicitly introduce the latinate terms as “foreign”. For example, when introducing Gegenstand and Äußerung as the basic parts of a sentence, he says: sie werden als solche mit einem fremden Namen das subject und prädicat genannt. ‘with a foreign name, they, as such, are called subject and predicate’ (Heyse 1827: 633). The grammar of Murray was similarly highy influential in the English speaking world, with numerous editions appearing throughout the 19th Century. However, he uses the terms “subject” and “object”, but never “predicate”, opting for “attribute” instead: “the principal parts of a simple sentence are, the subject, the attribute, and the object” (Murray 1824: 126).
[2.8] These are just a few examples of the gradual adoption of erstwhile logical terminology into the realm of grammatical description. The expansion of this terminology in the 17th and 18th Century needs a more detailed investigation than what I am able to provide here. But by the start of the 19th Century the terms subject, object, predicate and copula all seem to have been accepted as regular terms for grammatical analysis. Yet, there are two aspects of the logical origin of these terms that spill over into the grammatical realm and remain widespread in the 19th Century and beyond. These are, first, the balanced and privilaged bond between subject and predicate (cf. the “S-P-Funktion” in Forsgren 1992: 75-76) and, second, the universality of the copula (cf. the “Copula-Theorie” in Forsgren 1992: 135-153).
[2.9] There seem to be at least two reasons:
[3.1] The 19th Century is widely considered to be the heyday of German linguistics. Many of the great names from the history of linguistics were German scholars from this time period, like von Humboldt, Bopp, Grimm, Schleicher, Paul, and so on. Yet, none of these greats will be mentioned in this chapter. In contrast, most of the scholars that will be discussed here will be completely unfamiliar to most contemporeneous linguists. The main reason is that the well-known scholars from the 19th Century were not very much interested in syntactic analysis. In contrast, the more philologically and didactically interested linguists constantly were dealing with those long, complicated and even strange sentence structures that were produced by the well-respected and maybe even idolated authors and poets they were studying. To elucidate these long-winding sentences, either as part of their research into stilistic characteristics or to explain these structures to newer generations of philological students, many scholars started using graphical methods for syntactic analysis. Although many of the authors to be discussed in this chapter were pondering didactic consideration, there is also an important strain of stilistic research that was an important impetus for these developments.
[3.2] In German grammars of the 18th and 19th Century there is a widespread conceptual distinction between a bare and an extended sentence. A bare sentence is mostly defined as a sentence with only a subject and a predicate without any attributive or adverbial modifiers; a minimal sentence of sorts. The examples given always have intransitive verbs and it remains unclear whether transitive sentences also could be “bare”. An extended sentence, in contrast, contains additional modifiers, though subordinate clauses are mostly relegated to yet further categories of sentence structure. This distinction is mostly referred to using the German words nackt ‘naked’ vs. ausgebildet ‘developed’. However, there is some interesting variations and discussion throughout the 19th Century about exactly which word should be used.
[3.3] The origin of this terminological distinction can probably be traced back to Meiner (1781: 446), who uses the terms bloß ‘bare’ for a minimal sentence and ausgebildet ‘developed’ for a sentence with additional modifiers. Contemporaneously, Adelung (1782: 572) uses the terms einfach/nackt ‘simple/naked’ and ausgebildet ‘developed’. This terminology is slightly changed in the first edition of Heyse’s influential grammar of German (Heyse 1814: 492-492) into einfach/rein ‘simple/pure’ and ausgebildet/erweitert ‘developed/extended’. Then Heyse adjusted this again in the second edition by adding the term nackt ‘naked, bare’ as an alternative to einfach and rein (Heyse 1820: 598). Finally, in the fourth edition of 1827 Heyse adds the term bekleidet ‘dressed’:
Solche Sätze […] kann man ausgebildete, erweiterte, oder (im Gegensatze zu den nackten) bekleidete Sätze nennen. (Heyse 1827: 636)
(‘Such sentences can be called “developed”, “extended” or (opposite to “naked”) “dressed” sentences.’)
[3.4] The term bekleidet leads to a lot of discussion, because of the sexual connotations. The term nackt is not so much the problem, because this German word is used to refer to many situations without cover, like nackter Boden ‘bare floor’, and it is also regularly used metaphorically, like in nackte Angst ‘sheer terror’. So, there is no sexual connotation with the usage of the word nackt when used in opposition to the words ausgebildet or erweitert. However, in opposition to the word bekleidet ‘dressed’ the meaning of the word nackt suddenly obtains a strong sexual connotation, clearly meaning ‘nude’. Some authors (e.g. Billroth 1834; Götzinger 1835a) use bekleidet, maybe because they find it funny and consider it a nice joke to motivate students, while others strongly object to this choice of words as being vulgar, or at least tasteless.
Man nennt solche Sätze, welche nur die wesentlichen (Haupt-) Satztheile enthalten, nackte, und solche, welche auch noch andere Theile enthalten, (nicht: bekleidete, sondern) ausgebildete oder erweiterte Sätze. (Diesterweg 1830: 166)
(‘Sentences which contain only the essential (main) parts are called nackt ’naked’, and those which also contain other parts are called (not: bekleidet ‘dressed’, but) ausgebildet ‘developed’ or erweitert ‘expanded’ sentences.)
Von nackten und bekleideten Sätzen zu reden ist geschmacklos, von nackten und erweiterten ist nicht viel geschmackvoller und enthält außerdem einen schiefen Gegensatz, den man in keiner Schülerarbeit dulden würde. (Kern 1883a: 5)
(‘To speak of naked and dressed sentences is tasteless, to speak of naked and extended sentences is not much more tasteful and also contains a distorted contrast that would not be tolerated in any student’s work.’)
[3.5] The term Periode as a grammatical term for the complex sentence is widespread in German 18th and 19th Century grammatical texts. It is used already by Meiner (1781: 446) and Heyse (1814: 491) and appears in many works to be discussed in this chapter. The precise definition is constantly being disputed, but most authors use the term Periode for a complex sentence with many subordinated clauses.
[3.6] This term that goes straight back to antiquity. Aristotle used the term περίοδος ‘periodos’ to describe something akin to a complete sentence, consisting of multiple clauses (with κῶλον ‘kolon’ being used for something akin to a clause). Burkett explains that “the term periodos is one of Aristotle‘s visual metaphors, suggesting a circular motion or completeness of thought and rhythm.” (Burkett 2011: 206).
λέγω δὲ περίοδον λέξιν ἔχουσαν ἀρχὴν καὶ τελευτὴν αὐτὴν καθ᾽ αὑτὴν καὶ μέγεθος εὐσύνοπτον (Aristoteles, Rhetorica III 9.3) Original online at the Perseus Digital Library < http://data.perseus.org/citations/urn:cts:greekLit:tlg0086.tlg038.perseus-grc1:3.9.3>.
(‘I call a period an expression having a beginning and an end in itself and a magnitude easily taken in at a glance.’ translation from Burkett 2011: 206)
[3.7] The desire to summarise the structure of a complex Periode is a crucial starting point for the graphical analysis of sentences in 19th Century Germany, starting with Herling (1823, see Section 3.2.2). The main tool that is used is a so-called Periodenbild ‘image of a Period’. The basic idea is to assign some symbols to each clause, and then combine these symbols into a visualisation to illustrate the interrelation between the many clauses that make up a Periode. The term Periodenbild for the analysis of a multi-clause sentence was first used by August Lehmann (1833, see Section 3.2.8). His proposal includes a distinction between capital letters for main clauses and lower-case letters for subordinate clauses, a principle that remains a fixture throughout all variants of these Periodenbilder. In contrast, almost all other details differ widely from author to author.
[3.8] Friedrich Diesterweg (1834, see Section 3.2.6) adopted Lehmann’s terminology and added the term Satzbild for the structure of a monoclausal sentence. The two terms Periodenbild and Satzbild return throughout the 19th Century, but they are used quite interchangeably. Even though many authors distinguish between two different graphical analyses, one for single-clause sentences and one for multi-clause sentences, Diesterweg’s terminological distinction did not persist. Linguistic Bilder remain virulent for about 90 years, but then gradually fade away from linguistic memory at the start of the 20th Century.
[3.9] Karl Philipp Moritz (1756-1793) was a prolific author with wide ranging interests. Towards the end of his life he became professor for the arts at the Königliche Akademie der Künste in Berlin. Among his many writings there are various instructional books and articles about grammar, see Schmidt (1993) for a concise summary of his linguistic oeuvre. Of particular interest here is the Deutsche Sprachlehre für die Damen in Briefen (Moritz 1782). The context and linguistic content of this work is discussed in more detail by Buhofer (1994). This book was quite successful, being reprinted in 1791, 1794 and 1806. All later editions dropped the somewhat curious addition für die Damen from the title.
[3.10] The Deutsche Sprachlehre contains to my knowledge the first graphical display of a grammatical analysis. In the Zehnter Brief: Von dem Bau der Rede, starting on page 267, Moritz discusses various simple examples, culminating in two large fold-out page-inserts between pages 308 and 309, reproduced here in Figure 3.1 and Figure 3.2. Later, in the first volume of his 1793 Grammatisches Wörterbuch der deutschen Sprache there is a short summary of this grammatical analysis under the heading of Construktion (Moritz 1793: 283-295). After Moritz’ early demise that same year (at only 37 years of age) this Grammatisches Wörterbuch was completed by other authors without any further mentioning of this new kind of sentence analysis. The most extensive exposition of Moritz’ envisioned analysis thus remains the discussion in the 1782 Deutsche Sprachlehre.
[3.11] Moritz’ linguistic analysis is particularly noteworthy because it is an early version of what later will be called a Satzgliedanalyse (‘phrasal analysis’). Although Moritz consistently talks about the “words” of the sentence, he clearly intends these “words” to be something closer to what today would be called phrases. These phrases are classified into various functions, using German labels for didactic purposes (e.g. Grund=subject, Ziel=accusative object, Zweck=dative indirect object). Forsgren (1985: 76), discussing the origin of such phrasal analyses at the end of the 18th Century, argues that Moritz’ approach is one of the earliest examples of a phrasal analysis and his ideas appear to be developed independently from other contemporaries.
[3.12] Examining the details of this visual display, each clause can be identified by a large bracket. Within each clausal bracket there is a table-like structure with function-names on top and the actual words of the clause below. Each clause consists of Grund (“subject”) and Rede (“predicate”), with possibly additional syntactic functions like Ziel (“accusative object”), Zweck (“dative object”) or Bindung (“conjunction”). Optional elements like Hinanfügung (“adverbial phrase”) and Einschiebung (“interjection”) are placed loosely below the main syntactic functions, implicitly hinting at an argument/adjunct distinction. Various details are not completely thought through and the examples are sometimes slightly inconsistent. For example, the phrase von der Schulter in clause 11 is analysed as an adverbial Hinanfügung, while the completely parallel phrase auf die Schulter in clause 18 is analysed as an object-like zweites Ziel.
[3.13] More interesting is the treatment of subordinate clauses. Subordinate clauses are given a function in the matrix clause and then are further analysed within a new bracket as a separate clause below. This can be seen in Figure 3.2 with the analysis of direct speech in clause 13, which is the Ziel of the matrix clause 12. Likewise, the adverbial um zu+Infinitiv subordinate clause 21 is the Zweck of the matrix clause 20.
[3.14] This display of subordinate clauses is no coincidence. In the text there is an explicit discussion of this analysis of subordinate clauses (Moritz 1782: 299-300) based on the example sentence (3.1). This sentence is analysed by Moritz into four separate clauses. The four clauses and their functions in the superordinate clause are listed in (3.1 a-d), as shown in Figure 3.3. Each of these subordinate clauses is subsequently analysed using the same clausal analysis structure as the superordinate clause, as can be seen in Figure 3.4.
(3.1) | Wenn ich meine Pflicht thue, so fühle ich, daß mir dieses genug ist, um vergnügt und glücklich zu sein. | |
a. | Grund und Rede: so fühle ich | |
b. | Ziel: daß mir dieses genug ist | |
c. | Zweck: um vergnügt und glücklich zu sein | |
d. | Hinanfügung: wenn ich meine Pflicht thue |
[3.15] Nitpicking one might argue that this analysis of the complex sentence is not completely correct. Clause 2 from Figure 3.4 is not a separate clause but the superordinate clause itself. Further, The subordinate part um vergnügt und glücklich zu sein is analysed in Figure 3.3 as Zweck (‘reason’, confusingly a label otherwise also used for dative arguments by Moritz). However, this clause is not directly subordinate to the main clause so fühle ich but it is secondary subordinate to the Ziel complement clause daß mir dieses genug ist.
[3.16] Notwithstanding these inaccuracies, Moritz’ analysis shows a clear understanding of subordinate clauses having simultaneously an internal functional structure and, as a whole clause, having a function inside a superordinate structure. Anachronistically interpreting this as a constituency analysis, this sentence can be displayed as shown in Figure 3.5. Note that this is somewhat akin to an “underlying” structure as the ordering does not correspond to the original sentence. Moritz consistently applies the same functional structure on each clause (viz. the ordering Grund-Rede-Ziel-Zweck-Hinanfügung). The numbering in this example indicates the eventual ordering in the resulting sentence.
[3.17] Summarising, Moritz appears to have a clear understanding of, first, phrases as the basis of syntactic analysis. Second, he assigns a syntactic function to each phrase. Third, his examples indicate a basic, through rather implicit, understanding of adjuncts as less crucial phrases (Hinanfügung) that are visually displayed below the main constituents. Fourth, subordinate clauses have their own clausal structure, while the subordinate as a whole is assigned a function in the superordinate clause, which is a clear precursor of constituency structure. Finally, in the text he explicitly mentions that the ordering of the elements in the actual sentence is slightly different from the ordering in his analyses, and he hints at rules that determine how the abstract “underlying” structure has to be reordered into the actual form.
[3.18] All these theoretical concepts are quite charmingly summarised by his graphical display, which can be interpreted as a kind of precursor of a constituency structure. Apparently appearing completely out of the blue, the idea of using graphics to elucidate grammatical structure is off to a great start with these proposals from Moritz’ Deutsche Sprachlehre. Unfortunately, his approach does not have had any following and it would take almost 50 years for graphical methods to be used again for grammatical analysis, by Diesterweg in 1830 (see Section 3.2.6).
[3.19] Simon Heinrich Adolf Herling (1780-1849) was a teacher in Frankfurt am Main and founding member, and later main organiser, of the Frankfurter Gelehrtenverein für deutsche Sprache. His analysis of multi-clause sentences is widely considered foundational for the grammatical analysis of German. A wide-ranging appraisal of Herling’s work can be found in Elmentaler (1996). Herling is a close friend and colleague to Ferdinand Becker (see Section 3.2.5), who is also an active member of the Frankfurter Gelehrtenvereion. A plan to write a joint grammar fell apart, indicating different opinions in their approach to grammatical analysis (Haselbach 1966: 69). Their relation remains amicable, as indicated by the fact that the first part of Herling’s Syntax der deutschen Sprache (Herling 1830) is dedicated to Becker.
[3.20] Strictly speaking, the work of Herling does not include any graphical display of grammatical analysis. However, his work is still included here because he introduces some notation that will be expanded upon graphically by later authors, especially Lehmann (see Section 3.2.8) and Götzinger (see Section 3.3.1). Herling’s notational framework is first introduced in the Grundregeln des deutschen Stils oder der Periodenbau der deutschen Sprache. Ein Lehrbuch für den stilistischen Unterricht (Herling 1823), which later becomes the second part of Die Syntax der deutschen Sprache (second edition 1827, third edition 1832).
[3.21] In his book Herling analyses the different ways a Periode is constructed from multiple clauses. In a short note inserted in small font inside the main text, Herling proposes a special notation um die Umformungen kürzer zu bezeichnen, und so leichter überschauen zu können ‘to succinctly mark the transformations and thus be able to survey them more easily’ (Herling 1823: 177). The usage of the term Umformung deserves it’s own discussion, which will not be pursued here. The special notation is summarised slightly clearer in the second and third edition (e.g. Herling 1832: 232, see also 177). The version from the third edition is shown here in the Figure 3.6, which can be translated as follows:
For the succinct marking of the sentences and their syntactic relationships may the following symbols be of use: h indicates grammatical main sentence; b adjective clause, β, when it is shortened; a adverbial clause, α, the same shortened; s nominal clause, σ the same shortened. The exponents, e.g. a2 indicate the grammatical grade, to h e.g. h2 the grade of the logical gradation; h/2, h/3, indicates that the main clause, or a/2, a/3 that the adverbial clause is separated into two or three parts. The coefficients e.g. 2h indicate the number of equivalent main clauses; with a, b, s their relationship to the main clauses, often their coordination. (Herling 1832: 232)
[3.22] Herling does not use his notations consistently throughout his book. There are only a few incidental examples and there are almost no sentences completely marked according to his system. The example in (3.2) is one of the more interesting and complex cases (Herling 1832: 233). Yet, in contrast to intended purpose, the abstract analysis with letters and numbers, interspersed inside the sentence, does not really ease the quick appraisal of the sentence structure, in my opinion.
(3.2) | Um (α) dem Soldatenstande zu entgehen, gegen den (b2) er, ungeachtet (α3) seinen unternehmenden Geistes und kräftigen Körpers, entschiedene Abneigung fühlte, verdingte er sich, (1h) im sechszehnten Jahre, bei einem Gastwirthe an der Gränze als Fuhrknecht, und (2h) kehrte erst nach sieben Jahren, da (2a) eben in seinem Geburtsorte ein nicht ganz unbedeutendes Grundstück feil geboten ward, in die Heimath zurück. (Herling 1832: 233) |
[3.23] Anachronistically interpreting Herling’s analysis in modern terms, his symbols can be seen as a constituency structure. All clauses in the complex sentence are given a label and are related to each other in a hierarchical fashion, distinguishing coordination (viz. number in front) from subordination (viz. special letters for different kinds of embedding and superscript numbers to indicate deeper levels of embedding). Crucially, the clauses are linked to each other without using specific words as anchor points, like they would be in a dependency structure (cf. Becker’s sentence analysis of a complex sentence, see Section 3.2.5). Figure 3.7 shows an interpretation of Herling’s analysis from (3.2) in the form of a constituency tree.
[3.24] The analysis of the complex sentence remains a topic of Herling’s subsequent work. Noteworthy in the current context is a note to the teachers at the end of the Erster Cursus eines wissenschaftlichen Unterrichtes in der deutschen Sprache für Deutsche, nach einer neuen auf die Bildungsgesetze der Sprachen gegründeten Methode (Herling 1828). This book does not have any graphics in the main text. However, completely at the end, in the Appendix Bemerkungen und Erläuterungen für den Lehrer (Herling 1828: 311-367), he mentions the symbolic analysis using letters as discussed above (Herling 1828: 365-366). And then, as a cherry on top, he proposes to use the well-known graphical display of musical notation to illustrate the structure of complex sentences, as shown in Figure 3.8. Herling does not seem to have taken this idea further, only Diesterweg did (see Section 3.2.6). Unfortunately, this innovative proposal appears to have disappeared after that.
[3.25] The example in Figure 3.8 consists of two paragraphs, a “good” and a “faulty” example. The good example (3.3) is from the philosopher Johann Jakob Engel. It consists of five sentences, which are indicated in the notation by the (vertical) musical bar lines. The main clauses (h) are basically placed on the middle line, though some are on higher lines. It is not completely clear from the text how exactly this difference in “height” is determined. Subordinate clauses are written below the main clauses, depending on the depth of the embedding. There is a curious addition of “speed” in the graphical display, as indicated by the quarter and sixteenth notes. It seems to be used for stretches of parallel clauses.
(3.3) | So lange der Mensch nicht reden konnte, so sah, hörte, fühlte und schmeckte er bloß; aber er dachte nicht. So langue der Mensch nicht schreiben konnte, dachte er wenig, und redete schlecht. Die Zunge und der Griffel machten endlich den Menschen zu dem, was er werden sollte. Seine Begriffe wurden hell, indem er sie mitzutheilen suchte; sie wurden methodisch, indem er ihnen eine gewisse Fortdauer gab, die sie der Verbesserung, und Ausbildung fähig machte. Und dieser Weg, den das ganze menschliche Geschlecht nahm, um klüger zu werden, ist auch immer noch der einzige für den einzelnen Menschen. (Herling 1828: 305-306, uncredited quotation from J.J. Engel (1775) Der Philosoph für die Welt, Siebzehntes Stück Das Weihnachtsgeschenk) |
[3.26] Maximilian Wilhelm Götzinger (1799–1856) was a teacher of German, most of his life working in Schaffhausen (Switzerland). He wrote extensively about grammar, literature and dialectology, see Olsson (2009) for an comprehensive appraisal of his work. Besides his didactic writings, Götzinger also provided material for Grimm’s dictionary project (Olsson 2009: 37) and he was even awarded a doctorate honoris causa from the University of Basel in 1838 (Olsson 2009: 36), probably through his personal contact with Wilhelm Wackernagel (the father of Jacob Wackernagel of Wackernagel’s law-fame).
[3.27] Götzinger wrote two didactic grammatical textbooks, both of which appeared in many different editions throughout the 19th Century (see Olsson 2009: 42-46 for a list of all editions). The book Die Anfangsgründe der deutschen Sprachlehre in Regeln und Aufgaben was first published in Leipzig in 1825. This was followed in quick succession by the Deutsche Sprachlehre für Schulen, first published in 1827 in Aarau. He also wrote a much more interesting book in 1836, which will be discussed separately (see Section 3.3.1).
[3.28] There is a curious graphic in the first edition of the Deutsche Sprachlehre für Schulen (Götzinger 1827: 263), reproduced here in Figure 3.9 from the second edition because of the higher quality of the scan (Götzinger 1830: 263). The graphic indicates the Hebung ‘raising’ and Senkung ‘lowering’ of the voice at the various Glieder of a Periode, so this is not a graphical display of syntactic structure. For example, the example sentence in (3.4) illustrates the option Viergliedrig, in this case probably intended to show a case with three rising Glieder and a final falling one.
(3.4) | Wo göttliche Kraft und Empfindung die Gedanken belebt; wo Dank und Bitte auf den Flügeln des Wortes zum Himmel steigt; wo ein heiliger Ort den umherschweifenden Sinn feßelt und sammelt – da ist Religion. (Götzinger 1827: 264, citing an example from Köppen) |
[3.29] Friedrich Jakob Schmitthenner (1796-1850) was school director and later professor for history and organisation of state at the university of Gießen. He also wrote various books about German grammar and was part of the Frankfurtischer Gelehrtenverein für deutsche Sprache, alongside Herling (Section 3.2.2) and Becker (Section 3.2.5).
[3.30] In his 1828 book Teutonia. Ausführliche Sprachlehre nach neuer wissenschaftlicher Begründung. II. Buch. Höhere Sprachlehre Schmitthenner includes a graphic for the Periode (Figure 3.10). This graphic is very similar to Götzinger’s graphic (see Section 3.2.3), though it is unclear whether there is any direct influence. Like Götzinger, Schmitthenner explains that in his graphic the up and down strokes are supposed to indicate rising (Hebung) and falling (Senkung) intonation, and not syntactic structure. (Schmitthenner 1828: 185)
[3.31] Karl Ferdinand Becker (1775-1849) was probably one of the most influential grammatical educators from the start of the 19th Century. He originally started out as a medical doctor, but that turned out not be a financial success. Instead, he opened a small private school in Offenbach teaching German to English-speaking students. Turning his attention to linguistics, he wrote various books about German grammar. For a thorough repraisal of Becker’s linguistic work, see Haselbach (1966). Becker also became a member of the Frankfurter Gelehrtenverein für deutsche Sprache, where he met Herling (see Section 3.2.2) and Schmitthenner (see Section 3.2.4).
[3.32] In his Organism der Sprache als Einleitung zur deutschen Grammatik (Becker 1827) Becker presents a clear plan for analysis of syntactic relationships inside the sentence (which he called Satzverhältnisse), and it is easily possible to anachronistically depict these ideas graphically (e.g. Haselbach 1966: 156-161). However, Becker himself never produced any graphical display of the analysis of a specific sentence, but he came very close with a tabular method of analysis. He originaly introduced a few examples of his tabular method in the preface of the Deutsche Grammatik (Becker 1829). Here he presents a few examples of how sentences should be analysed, calling the process Analyse der Sätze (Becker 1829: xvi-xix). This explanation is addressed to teachers, instructing them how students should approach the analysis of a sentence. Astonishingly, he never describes this analytical approach in the grammar itself. He returns to this approach in a short book Über die Methode des Unterrichtes in der deutschen Sprache, again directed to teachers (Becker 1833: 36-46). In this book he calls the process Zergliederung ‘dissection’.
[3.33] Becker’s basic idea is to use a table with all words of the sentence listed as the rows of a table. This approach feels completely natural in the English grammatical tradition of “parsing”, which he might have encountered while teaching English-speaking students. In the English grammatical tradition, parsing has a long history. The English word itself originates from the Latin pars orationis, i.e. originally parsing is the process to assign parts of speech to individual words. The Oxford English Dictionary lists examples of this usage all the way back to the 16th Century. However, the process of consistently going through a sentence, listing all words separately, and analysing each one individually, seems to have been first used by John Brinsly the Elder in his Ludus Literarius from 1612 (nice examples in Brinsley 1612: 102, 130). In the centuries since, this approach to parsing became a mainstay in the English grammatical tradition (Lyman 1922: 120-122). Curiously, this approach to sentence analysis seems to have been basically unknown outside of the English-speaking grammatical tradition.
[3.34] An example of Becker’s table is shown in Figure 3.11, analysing the example sentence in (3.5). The crucial innovation of Becker is the addition of numbers to each row of the table, an approach never before, and never since, used in the English tradition. Becker uses these numbers to clarify the relationships between the words, as for each word he adds the number to the word it modifies at the end of each row in the table. An intriguing detail is that the predicate is linked towards the subject, suggesting that the predicate is modifying the subject. This is in line with the following statement of Becker: “Das Subjekt als derjenige Faktor, auf den das Prädikat bezogen ist” (Becker 1829: 290), but at odds with other quotations, like: “… den Begriff des Subjektes, der dem Hauptbegriffe untergeordnet ist, …” (Becker 1831: 15).
[3.35] Another aspect stressed by Becker is that the analysis has two parts: the Wortform (“parts of speech”) and the grammatische Bedeutung (“grammatical meaning”). This twofold analysis might very well be inspired by the highly influential English grammar of Murray, who likewise distinguishes two different kinds of parsing: “[parsing] should be divided into two parts: viz. parsing, as it respects etymology alone; and parsing, as it respects both etymology and syntax” (Murray 1824: 195).
Auch darf die Analyse sich nicht darauf beschränken, etwa nur die Wortarten zu unterscheiden: sie muß bei jedem Gliede des Satzes seine grammatische Bedeutung und die besondere Art der Beziehung bezeichnen, durch welche es mit dem Satze verbunden ist. (Becker 1829: xvi)
(‘The analysis should not be restricted to distinguish the parts of speech: it has to indicate for each part of the sentence its grammatical meaning and the special kind of relationship, with which it is connected to the sentence.’)
(3.5) | Wallenstein wußte längst den Inhalt ihrer Sendung als die Abgesandten des Kaisers ihm vor die Augen traten. |
[3.36] The format of the table is actually almost completely identical to the contemporary CoNLL-U table-format https://universaldependencies.org/format.html for dependency trees. In that format all words are listed in a tab-delimeted file, with minimally five columns: (i) a numerical ID, (ii) the linguistic element, (iii) the word class, (iv) the relation to the head, and (v) the number of the row of the head it modifies. The resulting dependency tree is shown in Figure 3.12. The file to reproduce this tree is linked at the end of the caption, and is almost verbatim the table from Becker’s book of 1829.
[3.37] A more extensive description of this tabular method is given in Becker (1833: 38-46), now using the Term Zergliederung ‘dissection’. From the exposition in that book the impression arises that Becker considered this kind of analysis the basis of all grammatical education. He stresses how important it is to practice the Zergliederung with students, and that all other aspects of the grammar can be better explained and understood when they are able to perform such a Zergliederung. Given this perspective, it remains completely unclear why this method is only presented in texts adressed to the teachers, and never explained to the students in the actual grammar books.
[3.38] An example of such a Zergliederung from Becker (1833: 43) is presented in Figure 3.13 for the sentence in (3.6). This example is interesting because it contains various subordinated clauses. Becker’s method is easily extended to deal with such complex sentences. I leave it as an exercise for the reader to turn this table into the CoNLL-U format and generate a dependency tree.
[3.39] In summary, the grammatical writings of Becker have had a lot of influence in the 19th Century, but his very practical idea of a numbered tabular analysis has unfortunately not been widely acknowledged. Diesterweg used it, but his proposals also did not have a large impact (see Section 3.2.6). Likewise Fürg adopted it (see Section 3.3.7), but that was the last time this method seem to have appeared in the literature. Still, Becker did have an influence on the graphical grammatical practice in Great Brittain, but in a slightly different way (see Section 4.3.1).
(3.6) | Jakob schickte seinen jüngsten Sohn nicht mit den übrigen nach Aegypten, weil schon einmal ein Sohn, den er ausgesandt hatte, nicht zurückgekehrt war, und weil er fürchtete, daß auch diesem ein Unfall begegnen möchte. |
[3.40] Friedrich Adolph Wilhelm Diesterweg (1790-1866) was a teacher and a prolific author of textbooks. He also wrote extensively about didactic methods in general, strongly influenced by the ideas of Pestalozzi. Later in life he was politically active in an attempt to reform the Prussian educational system. His practical textbooks mainly dealt with mathematics and German linguistics. A comprehensive analysis of his linguistic work can be found in Czoska (1984). His linguistic proposals did not have much influence on the further development of syntactic analysis, which is unfortunate, because there are a lot of useful insights hidden there.
[3.41] The Praktischer Lehrgang für den Unterricht in der deutschen Sprache is a massive three-volume guide for teachers. As it says in the subtitle, it is ein Leitfaden für Lehrer, welche die Muttersprache naturgemäß lehren wollen (‘a guide for teachers who want to teach the mother tongue in a natural way’). In the current context I will concentrate on the second volume Die Wortformen und die Satzlehre. This second volume first appeared in 1830 (Diesterweg 1830), but there have been some important changes in the second edition from 1834 (Diesterweg 1834). The third edition from 1838 (Diesterweg 1838) and the fourth from 1847 only saw minor changes. I will cite the 1830 and 1834 version where appropriate, but often include references to the 1838 version as well, as this is the highest-quality openly available digital version. For that reasons I have also used the 1838 for the images included here, as they do not change from the 1830 or 1834 versions.
[3.42] In the preface to the second edition from 1834, Diesterweg calls his approach Elementarmethode ‘elementary method’ (a clear nod to Pestalozzi), in opposition to the wissenschaftliche oder die constructive Methode ‘scientific or constructive method’ (Diesterweg 1834: xii-xiii). As examples of the scientific method he explicitly cites Herling, Becker, Götzinger and Schmitthenner, all of which he deems inappropriate for secondary education. However, the subsequent proposals Diesterweg makes in his book are all clearly inspired by these authors.
[3.43] In the first edition from 1830, Diesterweg includes a graphical represenation of the intonation contour of a Periode, following Götzinger (Section 3.2.3) and Schmitthenner (Section 3.2.4), shown here in Figure 3.14 (Diesterweg 1830: 191), shown here in the version from the third edition (Diesterweg 1838: 169). Diesterweg proposes a consistent uses of punctuation marks, with semicolons for sequences of Beiordung and colons for the main intonation maximum. With this analysis, Diesterweg follows the approach from Götzinger that each Periode has a single maximum.
[3.44] In the 1830 edition Diesterweg also proposes a notation for subordinated clauses (using a raised star: *) and coordinated clauses (using a the typographic dagger: †). Deeper embedded clauses are marked by repeating these symbols (Diesterweg 1830: 188-190). These symbols are placed at the start of each clause, resulting in examples like (3.7), somewhat reminiscent of the use of letters in (3.2) by Herling.
(3.7) | Nicht, * um dir einen neuen Beweis meiner zärtlichen Sehnsucht zu geben, *† sondern um feurige Kohlen auf dein feindseliges Haupt zu sammeln, schreibe ich dir noch eineml, † und zwar aus dieser berühmten Handelsstadt (Leipzig), * welche durch ihre Messen ganz Deutschland mit Waaren aller Art versorgt, *† und deren Name gleich berühmt ist durch die Leckerbissen, ** die sie den Reichen, *† wie durch Schätze der Gelehrsamkeit, ** die sie den Gebildeten liefert. (Diesterweg 1830: 189) |
[3.45] In the 1834 edition Diesterweg replaces these symbolic visualisations with letter sequences based on the proposals by Herling (see Section 3.2.2) and Lehmann (see Section 3.2.8). He introduces capital letters for main clauses and lower-case letters for subordinate clauses like Lehmann (1833) using the superscript numbers for ‘deeper’ subordinations like Herling Herling (1823). However, Diesterweg ignores the meaning of the different letters as proposed by Herling, similar to Lehmann (1833). Also, he ignores Lehmann’s use of horizontal braces, replacing them with simple commas. The resulting sequence of letters is called Satzbild by Diesterweg (1834: 162-163, 170), which is the first occurence of this term. However, at another page he uses the term Periodenbild with the same meaning (Diesterweg 1834: 168-169). This term was first used by Lehmann (1833).
[3.46] Additionally, Diesterweg combines the letter sequences with the graphical representation of the intonation contour (Diesterweg 1834: 162-163; Diesterweg 1838: 164-165). And as an alternative representation, Diesterweg also proposes to use the music-score diagram from Herling (1828). Both visualisations are shown in Figure 3.15 (Diesterweg 1834: 169; Diesterweg 1838: 171) illustrating the example sentence in (3.8). Just like Diesterweg, I leave it as an exercise to the reader to follow along with the analysis in the figure.
(3.8) | Wenn der Mensch allein auf der Erde steht, und alle Herzen, die ihn liebten, in ihr und unter ihr ausgeschlagen haben, und in dem Gewühle der Menge, die ohne Rast nach ihrem Ziele drängt, keiner bei ihm verweilen will; wenn sein Schmerz einsam auf seinem harten Lager stöhnt, und keine Stimme ihm antwortet, als der herzlose Wiederhall; wenn sein thränenloses, glühendes Auge umsonst ein anderes sucht, und seine starre Hand nach einer andern greift, und die Kälte der Menschen ihn noch empfindlicher quält, als die Kälte des Winters in seiner düsteren Kammer: so ist doch in diesem Jammer, den so Viele jammern, noch eines, welches die Dornen des Schmerzes zerdrückt; do ist doch ein Auge, das liebend auf ihm ruht, und eine Hand, die ihn von seinem Strohlager aufrichtet. (Diesterweg 1834: 169, citing an example from Friedrich Jacobs) |
[3.47] But Diesterweg is not yet finished. He also proposes a syntactic analysis based on the ideas of Becker (see Section 3.2.5). Diesterweg uses an intricate graphical analysis of clause internal structure for ausgebildete Sätze, as shown in Figure 3.16 (Diesterweg 1830: 174, 266; Diesterweg 1838: 158, 260). He uses the distinction between nackte ‘naked’ and ausgebildete ‘developed’ sentences here, which was widespread at the time. The question about the internal structure of a sentence only arises with ausgebildete ‘developed’ sentences, i.e. with sentences that have additional material other than a bare subject and predicate. In the figure, the ‘naked’ sentence, consisting of a bare subject and predicate, is put on top and the ‘development’ is added below.
[3.48] Diesterweg uses braces, both horizontal and vertical, to indicate groups of words in developed sentences. It is tempting to compare this graphical approach using braces with the analysis of Moritz (see Section 3.2.1). However, there is no indication that there has been any direct influence of Moritz on Diesterweg. It seems more plausible that the use of braces is a parallel development, because large braces is simply one of the few graphical tricks that were available for typesetting at the time.
[3.49] Diesterweg does not employ syntactic diagrams with lines, because he restricts line diagrams to the illustration of hierarchical subdivision of grammatical categories, e.g differentiating kinds of plurals (Diesterweg 1838: 8), kinds of nouns (Diesterweg 1838: 22-23) or kinds of adjectives (Diesterweg 1838: 53), and many more such examples through the book. In contrast, a sentence is not an example of hierarchical subdivision, but an example of hierarchical modification, which Diesterweg calls Bestimmung. It seems plausible that Diesterweg quite consciously avoided the usage of the same visualisation for these two different applications.
[3.50] The visual analysis with braces, which Diesterweg calls Zergliederung and sometimes Analyse (following Becker 1829: 16) includes various crucial syntactic details. Many of these details are not explained; Diesterweg simply claims that well-taught pupils should be able to follow along with the analysis of the complex sentence (3.9) as illustrated with the graphic in Figure 3.16.
Sind die Schüler bis hieher [sic] gründlich unterrichtet worden; so müssen sie im Stande sein, den nachfolgenden Satz, wie folgt, zu zergliedern und wieder zusammenzusetzen. (Diesterweg 1830: 174; 1838: 158)
(‘When the pupils have been taught thoroughly up to here, then they should be able to parse and assemble again the following sentence, as follows.’)
(3.9) | Ein unter den schwierigen Umständen standhaft gebliebener Vater von sechs unmündigen, in weiter Entfernung lebenden Kindern hat sich auch in diesem theuern Jahre durch rastlose Thätigkeit bei Tag und Nacht vor Tausenden seiner Mitbürger rühmlich ausgezeichnet. |
[3.51] Diesterweg’s visual analysis of (3.9), as shown in Figure 3.16, contains various important syntactic details that have to be inferred from closely inspecting the display:
[3.52] Diesterweg returns to this syntactic analysis at the end of the book (Diesterweg 1830: part 3, chapter nine; Diesterweg 1834: part 3, chapter ten) under the heading Die grammatische Zergliederung der Sätze (Diesterweg 1830: 263). He first repeats the previously used analysis using braces, but then turns his attention to sentences containing (multiple) subordinate clauses. Without any comment he omits the bracketed analysis and turns to a table-style analysis with numbered words and cross-references, exactly like Becker (see Section 3.2.5). Diesterweg probably noticed that the bracketed visualisation becomes exceedingly complex and confusing with more elaborated sentences. However, from his seemless switch between bracketed visualisation and numbered tables it becomes clear that these two approaches are interchangeable in his mind. It is just a matter of practical applicability which system to use in a particular example.
[3.53] Finally, there are two interesting terminological issues addressed by Diesterweg. First, he prefers the term Ausbildung over the widespread term Erweiterung, which is actually still an extremely common grammatical term in Germany today.
Man hat die ausbildung des Satzes auch erweiterung desselben genannt. Es ist Erweiterung, in so fern mehr Wörter hinzutreten, als der Satz länger wird. Es ist dies eine grammatische Erweiterung. Da aber die näher bestimmten Wörter eine engere Begriffssphäre erhalten, so ist es eine logische Verengerung. Deshalb wird das Wort erweiterung besser vermieden. (Diesterweg 1830: 167; 1834: 151)
(‘The development (Ausbildung) of a sentence is also called expansion (Erweiterung). It is an expansion because more words are added, and the sentence will become longer. So it is a grammatical expansion. However, because the modified words obtain a more narrow meaning, it is a logical narrowing. Therefore, the word expansion (Erweiterung) is better avoided.’)
[3.54] Second, throughout the discussion of syntax, Diesterweg frequently uses the nomen agentis Bestimmer ‘determiner’ for a word that modifies another. This appears to be a quite practical term, but unfortunately this terminological quirk has not caught on.
Die Ausbildung geschieht durch nähere Bestimmung der einzelnen Satztheile. Wir wollen diese Wörter, welche die Hauptsatztheile näher bestimmen, überhaupt bestimmer nennen. (Diesterweg 1830: 167; 1834: 151)
(‘The development (Ausbildung) happens by modification (Bestimmung) of the individual parts of the clause. In general, we will call such words, that further modify parts of the main clause, modifiers (Bestimmer).’)
[3.55] In summary, the grammatical analysis of Diesterweg is a complex tour-de-force, combining insights from various scholars of the preceding decade, synthesising everything into a coherent and visually appealing package. Additionally, he attempts to make the grammatical analysis practically useful for application in the classroom. Many ideas proposed in his book feel rather modern and might have kickstarted a fruitful dicussion about syntactic analysis. Unfortunately, Diesterweg’s Lehrgang does not seem to have had much impact on the further development of linguistics in general, nor specifically on the establishment of linguistic education. From here on, the development of grammatical analysis in Germany will proceed taking baby steps, instead of leaping ahead like Diesterweg demonstrated to be possible.
[3.56] At the end of his short life, Johann Gustav Willhelm Billroth (1808-1836) became professor for religion at the university in Halle. Before that, between his Habilitation in Leipzig in 1830 and the start of his position in Halle in 1834, he worked as a teacher for latin (Alwast 2000). In that context he wrote two latin grammars for use in schools, first the Lateinische Syntax für die obern Klassen gelehrter Schulen (Billroth 1832), later followed by the Lateinische Schulgrammatik für alle Klassen (Billroth 1834).
[3.57] In these grammars there is a single graphical display of a grammatical structure, shown below in Figure 3.17 (Billroth 1832: 102). There is not much known about the grammatical background of Billroth, nor is there any explicit indication in his books about the inspiration for drawing a grammatical analysis. As he was active as a school teachter, he might very well have been aware of Becker (1829, see Section 3.2.5) or Diesterweg (1830, see Section 3.2.6), both of which might be inspiration for Billroth’s graphical display. One (minor) argument that Billroth was inspired by Diesterweg is that Billroth uses the term bestimmter Satz as an alternative to bekleideter Satz (Billroth 1832: 97), and the noun Bestimmung in the meaning of modifier (Billroth 1832: 126-127). Diesterweg very extensively used the nound Bestimmung but also Bestimmer, which is not found in Billroth. It is possible that Billroth just used these words out of his own inspiration, but the usage is idiosyncrativ enough to be notewhorthy.
[3.58] In Billroth’s syntax, the whole discussion of the structure of the sentence takes 11 pages (Billroth 1832: 91-102). The structure of the text is really difficult to follow, although on closer inspection there is a clear, but drawn-out organisation. First, he introduces the simple ‘naked’ sentence as option A:
Ein Satz ist entweder: A. ein einfacher (oder nackter) d.h. ein solcher, welcher bloß aus Subject, Prädicat und Copula besteht. (Billroth 1832: 91)
(‘A sentence is either: A. a simple (or naked) one i.e. one such that it only consists of subject, predicate and copula.’)
[3.59] It takes 8 pages of explanations of the simple sentence until finally the expected follow-up to the option A occurs. Without clear formatting, the option B is announced innocuously on page 97, using a rather strange sentence structure that only makes sense when it is read immediately after the option A as quoted above.
B. oder ein bestimmter (bekleideter) d.h. ein solcher, wo entweder ein einzelner Theil oder mehrere durch Zusätze näher bestimmt werden. (Billroth 1832: 97)
(‘B. or a determined (dressed) one, i.e. one such, in which either a single part or multiple parts are further determined by additions.’)
[3.60] Then it takes another 5 pages of explanations of different kinds of such “dressed” sentences until finally this topic is closed with the conclusion quoted below. Billroth simply states (without explanation) that the example sentence (3.10) is ‘dressed’ as illustrated in Figure 3.17. In this quote, Billroth takes the metaphor of the naked/dressed sentence even one step further by using the term Körper ‘body’ for the naked sentence.
Auf diese Weise wird z.B. folgender Satz (3.10), dessen Körper eigentlich nur aus dem Subj.: Miltiades und dem Prädicat (mit dem Copula) reddidit besteht, bekleidet: (Billroth 1832: 102)
(‘In this way, the following sentence, whose body only consists of the subject Miltiades and the predicate (with copula) reddidit, is dressed:’)
(3.10) | Miltiades, dux Atheniensium, toti Graeciae libertatem paene oppressam in pugna apud Marathonem reddidit. |
[3.61] Interpreting his graphical illustration in more detail, the naked sentence is depicted by a horizonal line, while the additional elements of the “dressed” sentence are all shown as vertical/slanted lines. Billroth very clearly intends the vertical lines to represent some kind of dependency, with multiple levels of modification being illustrated by subsequent vertical lines. The structure of the graphic is quite similar to Diesterwegs proposal in Figure 3.16, but with lines instead of curly braces.
[3.62] Exactly the same approach with the same picture is found in his extended grammar written two years later (Billroth 1834: 332-333). After Billroth’s early death this grammar has seen various new editions. The second edition of 1838 is prepared by Friedrich Ellendt and uses exactly the same figure (Billroth 1838: 329). However, in the third edition, the picture is changed, erroneously, and has become completely incomprehensible, cf. Figure 3.18 (Billroth 1848: 358). Ellendt possibly did not completely understand what the figure is supposed to represent.
[3.63] Billroth’s grammar has an interesting afterlife, but unfortunately without any further development of his graphical innovation. Friedrich Ellendt writes a different grammer for younger students based on Billroth’s model, called Lateinische Grammatik für die unteren Klassen der Gymnasien. The discussion of bekleidete oder erweitete Sätze is very similar to Billroth’s, but it has different examples and it does not have a picture anymore (Ellendt 1838: 115-116). After the death of Ellendt, that grammar is in turn adapted by Moritz Seyffert as Dr. Friedrich Ellendt’s Lateinische Grammatik für die unteren und mittleren Klassen der höheren Unterrichtsanstalten. This grammar has seen very many revisions, editions and reprints under the heading Ellendt-Seyfferts lateinische Grammatik far into the 20th Century. However, Seyffert was not very impressed by the syntactical analysis of Billroth (as noted in the preface of his revision), and he rewrote the discussion of syntax wholesale.
[3.64] On a more positive note, Billroth’s grammar had a direct influence on Julius Hoffmann (see Section 3.3.3), and Hoffmann is quoted indirectly by Franz Kern (see Section 3.3.24). So it seems very plausible that Billroth’s line diagram was a seed that grew slowly but surely.
[3.65] Johann August Otto Leopold Lehmann (1802-1883) worked as a teacher in Danzig (Gdańsk) and later as a school principal in Marienwerder (Kwidzyn). In this context, he prepared various educational textbooks, mainly consisting of literary commentaries, but also including songbooks, dialect analyses and some works on local history. Early in his career he wrote a linguistic treatise about the structure of the complex sentence: Allgemeiner Mechanismus des Periodenbaues, nebst einem Versuche, an ihn eine Kritik der deutschen Periode anzuknüpfen (Lehmann 1833). His approach to the analysis of complex sentences is based on the work of Herling (see Section 3.2.2), who is explicitly cited as a source of inspiration (Lehmann 1833: xv-xvi).
[3.66] In the title of his book, Lehmann uses the term Periode, which is a term also used by Herling. This term stems from the domain of rhetoric and roughly indicates a coherent and well-structured utterance. Continuing the tradition of Herling, Lehmann uses the term Periode to designate a complex sentence that consists of multiple clauses. Herling introduced symbolic abbreviations for different kinds of clauses (see Section 3.2.2), but Lehmann takes this a step further and proposes a comprehensive visualization of the structure of a complex sentence, which he calls a Periodenbild.
[Die] innere Anschauung der Periode soll als auf äußere Anschauung begründet dargestellt werden. Die äußere Anschauung aber findet ihr Objekt in Bildern, welche das Skelet der Perioden wiedergeben. Ich nenne dies Periodenskelet anschauungsbild der periode oder periodenbild schlechtweg. (Lehmann 1833: xiv)
(‘The internal structure of the Periode will be depicted based on the outward appearance. The outward appearance shows itself as a picture, which reflects the skeleton of the Periode. I simply call this Periode-skeleton an image of the Periode or a Periodenbild.’)
[3.67] For the visualization of a Periodenbild Lehmann distinguishes three different kinds of subordination, indicated by different symbolic devices as shown in Figure 3.19 (Lehmann 1833: 58). These three notations basically indicate a subordinate position after (1), before (2) and internal (3) to the superordinate clause, respectively. More precisely,
[3.68] At first glance it seems curious that there are completely different notation for the placement in front (colon) and the placement behind (horizontal brace). However, there is clear rationale for this difference arising from the structural possibilities of the German language (although this rationale is not explained by Lehmann in his book). The German main clause has a special initial position (today mostly called Vorfeld) that can contain a single subordinate clause. So each main clause can only have a single Vordersatz, which Lehmann indicates with the colon. Internal and final subordination can be recursively repeated.
[3.69] Lehmann’s book completely focusses on the analysis of complex sentences into their constituting clauses, using his graphical display to summarize the relationships between the clauses. There are hundreds of examples of different kinds of complex sentences discussed in his book, but there is not a single analysis of a clause-internal structure at all. He appears to be much more interested in the number of possible schemas that could exist, which are all Umformungen of a complex sentence. For example, he discusses an example of a particular schematic analysis (apparently without any real linguistic sentence in mind) that could theoretically have 37,632 Umformungen (Lehmann 1833: 116-118). Lehmann immediately notes “Es wäre thöricht, bei dieser oder ähnlichen Perioden, die zu den Verwickeltsten gehören, all möglichen Veränderungen darzustellen.” (‘It would be foolish for such or similar Perioden, that belong to the most complex ones, to illustrate all possible changes.’ Lehmann 1833: 116-118).
[3.70] Clauses are indicated by using letters, an approach that was also used by Herling (cf. Section 3.2.2). However, Lehmann has a completely different approach to the use of the letters. Lehmann is only interested in the relative ordering of the subordinated clauses and he completely ignores what kind of subordination is used. In contrast, Herling uses different letters for different grammatical kinds of subordinated clauses, for example a for adverbial clauses. When there are three adverbial clauses in a sentence, then they are all called a by Herling. The letters are thus indications of the grammatical kind of the clause. Lehmann instead uses a different letter for each subordinate clause, giving each clause an individual label, but omitting any indication of the internal structure.
[3.71] This difference between Herling and Lehmann also concerns the use of Greek letters. Herling uses greek letters for non-finite subordinate clauses (using zu-infinitive or participial verb forms in German), but Lehmann does not consider non-finite clauses at all. He uses greek letters for doubly-embedded subordinate clauses. In contrast, Herling uses numerical superscripts for multiple grades of embedding. Lehmann uses different kinds of scripts to distinguish multiple grades of embedding. He uses capital letters for main clauses, lower-case letters for subordinate clauses and greek letters for doubly-embedded clauses. Third-grade embedding uses Fraktur capital letters (e.g. 𝔄, 𝔅) and fourth-grade embedding uses Fraktur lower-case letter (e.g. 𝔞, 𝔟). There is even a short discussion of a fifth-grade embedding, which is indicated with numerals (Lehmann 1833: 232).
[3.72] In contrast to Herling, Lehmann’s approach thus ignores many apects of the grammatical analysis of a sentence. However, his focus on the relation between clauses allows him in return to tackle really complex sentences. Consider, for example, the long and complex sentence in (3.11), which is a sentence from Immanuel Kant’s Anthropologie in pragmatischer Hinsicht (Lehmann 1833: 125-126). The internal structure of this sentence is summarized by Lehmann with a relatively simple Periodenbild as shown in Figure 3.20.
(3.11) | a Daß ferner, ɑ wenn es (das Kind) nun zu sprechen versucht, a das Radbrechen der Wörter es für Mütter und Ammen so liebenswürdig macht, es beständig zu herzen und zu küssen und es auch wohl durch Erfüllung alles Wunsches und Willens zum kleinen Befehlshaber zu verziehn: A diese Liebenswürdigkeit des Geschöpfs im Zeitraum seiner Entwicklung zur Menschheit, muß wohl auf Rechnung seiner Unschuld und Offenheit aller seiner noch fehlerhaften Aeußerungen, b wobei noch kein Hehl und nichts Arges ist, A einerseits, andrerseits aber auf den natürlichen Hang der Ammen zum Wohlthun an einem Geschöpf, c welches einschmeichelnd sich der Willkür eines andern überläßt, A geschrieben werden, d da ihm eine Spielzeit eingewilligt wird, die glücklichste unter allen, β wobei der Erzieher dadurch, 𝔄 daß er sich selber gleichsam zum Kinde macht, β diese Annehmlichkeit nochmals genießt. (Lehmann 1833: 125-126, citing Immanuel Kant Anthropologie in pragmatischer Hinsicht 1798: 5) |
[3.73] The different clauses in Figure 3.20 are listed below, exactly as it is explained in the text by Lehmann:
[3.74] Although Lehmann’s book does not appear to have been reprinted after the first edition from 1833, it has influenced later more widely read works by Götzinger (see Section 3.3.1) and Bauer (see Section 3.3.8), whose graphical displays will be reprinted up to 1935. Götzinger’s acknowledgement indicates that Lehmann’s approach might have had some real influence in didactic circles, although it has not endured to the present day:
Die von Herrn Director Lehmann in Marienwerden zuerst aufgestellten Periodenbilder haben bei vielen Schulmännern Beifall und Anwendung gefunden, und es macht mir große Freude, durch meine Lehrbücher zur Verbreitung dieser Methode, deren Verdienst ganz allein Herrn Lehmann gebührt, beigetragen zu haben. (Götzinger 1842: x, Vorrede zur fünften Auflage)
(‘The Periodenbilder, first drawn by director Lehmann from Marienwerder, have been praised and employed by many Schoolmen, and I am delighted through my textbooks to have contributed to the wider circulation of this method, the credit to which completely befits Mr. Lehmann.’)
[3.75] Lehmann himself remained active and used his system in various later publications. First, he used it in Goethe’s Sprache und ihr Geist (Lehmann 1852: 152-180) to analyse sentences by Goethe, which includes very complex examples of his notation. Surprisingly, Lehmann is regularly quite critical of Goethe and uses his notation to argue for the imbalanced sentence structure, for example (3.12) as analysed in Figure 3.21.
Schon das Bilde dieser Periode zeigt deren Uebelstände. Erstlich die große Masse der Sätze, zweitens der Mangel an Ebenmaß theils zwischen den koordinirten Hauptsätzen theils zwischen den koordinirten Nebensätzen, drittens die Vorherrschaft des Anfügungsgesetzes. (Lehmann 1852: 171-172)
(‘Alreade the image of this Periode shows its foul nature. First, the large number of sentences, second the lack of balance, partly beteen the coordinated main clauses, partly between the coordinated subordinated clauses, third the dominance of the Anfügungsgesetz’)
(3.12) | Wilhelm hatte während der Zeit seiner Regie das ganze Geschäft mit einer gewissen Freiheit und Liberalität behandelt, vorzüglich auf die Sache gesehen und besonders bei Kleidungen, Dekorationen und Requisiten alles reichlich und anständig angeschafft, auch, um den guten Willen der Leute zu erhalten, ihrem Eigennutze geschmeichelt, da er ihnen durch edlere Motive nicht beikommen konnte; und er fand sich hierzu um so mehr berechtigt, als Serlo selbst keine Ansprüche machte, ein genauer Wirt zu sein, den Glanz seines Theaters gerne loben hörte und zufrieden war, wenn Aurelie, welche die ganze Haushaltung führte, nach Abzug aller Kosten versicherte, daß sie keine Schulden habe, und noch soviel hergab, als nötig war, die Schulden abzutragen, die Serlo unterdessen durch außerordentliche Freigebigkeit gegen seine Schönen und sonst etwa auf sich geladen haben mochte. (J.W. Goethe Wilhelm Meisters Lehrjahre) |
[3.76] Later Lehmann wrote Sprachliche Studien über das Nibelungenlied (Lehmann 1856; Lehmann 1857) in which he again uses his system to analyse classical German literature. In the introduction he cites Nägelsbach (Lehmann 1856: 4) and Götzinger (Lehmann 1856: 7) and follows Nägelsbach simplification (see Section 5.1.1) to use a comma instead of the horizontal brace (Lehmann 1857: 8). With this simplification is becomes possible to write Periodenbilder inside the normal text of a book, because the Periodenbild is just a linear line of characters.
[3.77] Finally, Lehmann wrote a literary analysis Forschungen über Lessings Sprache (Lehmann 1875). In this book he again used his Periodenbilder, but just as one of the possible ways to elucidate structures in the literary oevre of Lessing (Lehmann 1875: 130-149). Interestingly, he ignores the simplification he used in his previous word and reverts back to his original notation from 1833.
[3.78] Lehmann’s examples of the application of the scheme to actual literary texts inspired many philological analyses in the decades to come. For example, via the Repetitorium of Menge (see Section 5.1.2) there is a direct line to the 1969 dissertation on the Amadisroman from Langholf (see Section 3.3.46) using similar graphical analyses as Lehmann, more than a century after the original proposals.
[3.79] Maximilian Wilhelm Götzinger (1799–1856) was a teacher of German, most of his life working in Schaffhausen (Switzerland). He wrote extensively about grammar, literature and dialectology, see Olsson (2009) for an comprehensive appraisal of his work. Besides his didactic writings, Götzinger also provided material for Grimm’s dictionary project (Olsson 2009: 37) and he was even awarded a doctorate honoris causa from the University of Basel in 1838 (Olsson 2009: 36), probably through his personal contact with Wilhelm Wackernagel (the father of Jacob Wackernagel of Wackernagel’s law-fame).
[3.80] Götzinger’s early textbooks were already mentioned in Section 3.2.3, but these early books did not contain any graphical analyses of sentences. But then Götzinger discovered Lehmann’s Periodenbau (1833) and that changed everything. In 1835 Götzinger wrote a review of Lehmann’ book (Götzinger 1835b) and in the third edition of the Sprachlehre from 1835 he introduced Lehmann’s graphical displays. Götzinger cites the influence of Lehmann in the preface (Götzinger 1835a: xiii-xiv) and he applies Lehmann’s graphical analysis both to clause structure, under heading Stellungen der Bekleidung (Götzinger 1835a: 243-248), and to sentence structure, under the heading Stellung der Nebensätze (Götzinger 1835a: 310-315). Götzinger distinguishes these two applications of Lehmann’s analysis by using different names: Satzbild for clause structure and Periodenbild for larger sentence structures consisting of multiple clauses (Götzinger 1835a: 311).
[3.81] Contamporaneously to revising this textbook, Götzinger prepared his magnum opus Die deutsche Sprache und ihre Literatur (volume one 1836; volume two 1839). This book also includes many graphical analyses inspired by the Periodenbilder from Lehmann (see Section 3.2.8). In this book, Götzinger does not explicitly cite Lehmann as his inspiration for the graphical display, but he mentions Lehmann’s vortreffliches Buch (‘excellent book’) in a different context (Götzinger 1839: 217). I will use these two volumes of Die deutsche Sprache to summarise Götzinger’s interpretation of Lehmann’s approach.
[3.82] There are two main differences between Lehmann and Götzinger. First, Götzinger simplifies Lehmann’s quite elaborate use of different character set. Lehmann uses latin, greek, and fraktur letters, in upper- and lower-case, with different meanings. Götzinger only uses latin capitals for main clauses and latin lower-case letters for subordinate clauses. Second, Lehmann originally used his graphical method only for main and subordinated clauses (Periode), but Götzinger argues that the same principle of subordination (Unterordnung) can be used for other kinds of linguistic structure as well. Consequently, Götzinger uses Lehmann’s notation also for (i) syllabic structure, (ii) clause-internal structure and (iii) clause subordination.
Auf diese Weiße ordnet die Sprache bei allen Vorgängen den jedesmaligen Stoff so, daß ein Theil desselben als untergeordnet erscheint, in der Silbe die Laute, im Worte die Silben, im Satze die Worte, in der Periode die Sätze. (Götzinger 1836: 15)
(‘Accordingly, language arranges the particular content in all occurrences in such a way that one part of it appears as subordinated: with syllables the sounds, with words the syllables, with clauses the words, with sentences the clauses.’)
[3.83] First, for the analysis of the syllable, Götzinger (1836: 221-224) uses the capital letter for the vowel as the centre of the syllable (Träger und Mittelpunkt der Silbe) and lower-case letters for the consontants as dependent on the vowel (erscheint stets als untergeordnet). As long as the syllabic structure is according to the sonority hierarchy (musikalisches Gesetz), it can be described as (multiple) prefixal and suffixal subordinations, as illustrated in Figure 3.22. Extrasyllabic initial s and affricate codas ps, tsch and chs are illustrated with different structures not shown here (Götzinger 1836: 224).
In der Regel findet bei der Gliederung der Silbe durch fortschreitende Unterordnung der Elemente ein Schwellen und Sinken, ein Steigen und Fallen, ein Crescendo und Decrescendo statt; d.h. die Silbe fängt mit dem lautlosesten Element an, geht zu dem tönenenden über, gelangt zur eigentlichen Stimme, und geht nun denselben Weg wieder abwärts, so daß mithin in jeder auf diese Weiße gebaute Silbe ein völlig musikalisisches Gesetz statt findet. (Götzinger 1836: 222)
(‘Normally, the arrangement of the syllable consists, by repeated subordination of the elements, of swelling and sinking, of rising and falling, of crescendo and decrescendo; i.e. the syllable starts with the most silent element, proceeds to the more sounding one, reaches the principal voice, and then follows the same path downwards again in such a way that each such constructed syllable obeys a completely musical law.’)
[3.84] Second, for the analysis of the internal structure of a clause Götzinger proposes to use the same notation (Götzinger 1839: 159-162), which he calls Satzbilder (Götzinger 1839: 161). A few examples are shown in Figure 3.23. The capital letter now refers to the verb and the lower-case letters refer to the constituents. Götzinger calls the lower-cased elements Bekleidung. Today a German grammar would call them Satzglied. The German main clause structure with two verb positions is indicated by a repeated capital letter, called trennbare Zusammensetzung by Götzinger (1839: 160). Similar terminology can still be found in contemporaneous German grammars.
[3.85] Relative to the main verb, the three possible positions of the Bekleidung (i.e. before, in between and after the verb) are called Voraussendung, Anfügung, Einschiebung, respectively (Götzinger 1839: 160). They are indicated with a colon, a horizontal brace and round brackets in the tradition of Lehmann (see Section 3.2.8), who calls them Vorausschickung, Anfügung and Einschaltung. Today a German grammar would say Vorfeld, Nachfeld and Mittelfeld. Götzinger also observes one of the main generalisation of German main clause structure, namely that the Voraussendung/Vorfeld never consists of more than one Bekleidung/Satzglied.
Mag also die Stellung der Bekleidung in jenem Satze seyn, wie sie will, immer wird nur das Bild entstehen [mit einer Voraussendung]. Niemals kann die Folge vorkommen [mit mehreren Voraussendungen]. (Götzinger 1839: 162)
(‘Although the position of the constituents in the sentence can be as it chooses, the resulting image will always have a single Vorraussendung. The ordering with multiple Voraussendungen cannot occur.’)
[3.86] Third, for the analysis of the multi-clause sentence Götzinger again uses the same notation. He separates the analysis of basic subordination (Götzinger 1839: 364-367) from the analysis of the Periode (Götzinger 1839: 374-403). This distinction is not very clear, as he comments that der Begriff periode ist etwas schwankender Natur ‘the concept of the Periode has somewhat of a wavering nature’. He introduces the equal sign for coordination (Götzinger 1839: 366). A slighly complex Periodenbild is shown in Figure 3.24 for an example sentence (3.13) from Goethe.
(3.13) | Wenn ich mich in einer mittleren oder großen Stadt umsehe und bemerke, wo denn die Menschen sich hinwenden, um ihren Abend zuzubringen: so findet sich immer, daß man dahin gehe, wo man grüßend begrüßt wird; wo man gerne hört und gehört wird; wo man beim gefelligen Gespräch und Spiel immer gewiß ist, seine Parthie zu finden. (Götzinger 1839: 391, citing Goethe) |
[3.87] Franz Cornelius Honcamp (1805-1866) was a teacher in Büren (Westfalen). He wrote various grammatical textbooks, strongly influenced by Becker (see Section 3.2.5) and Diesterweg (see Section 3.2.6). Because of a small article, in which he implicitly criticised the role of the church in schools, he was eventualy laid off by the Prussian state in 1854 and worked as a private teacher for the last few years of his life. More biographical details and a detailed discussion of the unfortunate clash with church and state is provided by Diesterweg (1866). A useful summary of Honcamps life and work is available online at https://www.lexikon-westfaelischer-autorinnen-und-autoren.de/autoren/honcamp-franz-cornelius, accessed 11 February 2025.
[3.88] Honcamp’s first grammatical book is called Leitfaden für die Sprachbildung in deutschen Volkschulen (Honcamp 1838). In the subtitle he mentions the explicit influence from the work of Becker: Begründet auf Dr. K.F.Becker’s grammatische Ansicht, und fachgemäß methodologisch angeordnet ‘based on Dr. K.F.Becker’s grammatical view, and properly methodologically arranged’. Specifically, Honcamp uses the tabular analysis from Becker, in which each row is numbered (Honcamp 1838: 259-260, 271-275). An extensive explanation of Honcamp’s syntactic analysis is provided in his Vollständige Anleitung zum Elementar-Unterrichte in der Sprachlehre (Honcamp 1841: 13, 21, 30, 87, 90, 224-226). This book is dedicated to Becker and follows his proposals almost verbatim. An example of a Becker-style numbered table from this book is shown in Figure 3.25 for the biblical example in (3.14). This numbering allows for concrete reference of dependent words to their heads. Exactly the same use of numbered reference is (again) used by contemporary dependency grammar formats like Universal Dependencies (UD). Similar examples of syntactic analysis are also included in later books by Honcamp, e.g. in his Gedanken über den Unterricht in der Sprachlehre (Honcamp 1845: 29, dedicated to Diesterweg) and Lehr- und Übungsbuch für den Sprachunterricht in niederdeutschen Landschulen (Honcamp 1851: 19-20).
(3.14) | Als in Kanaan die Theurung ausgebrochen war schickte Jakob zehn seiner Söhne nach Aegypten, weil er hörte, daß man dort noch Getreide kaufen könnte. |
[3.89] Karl August Julius Hoffmann (1812-1869) was a schooldirector in Lüneburg. He wrote about such disparate topics as Homer, rhetoric, logic, and German grammar. In 1839 he published the first edition of his Neuhochdeutsche Schulgrammatik (Hoffmann 1839), which he revised in the second edition (Hoffmann 1856), but which did not get reprinted after that. In parallel, he also delivered a summarised version of the same work, the Neuhochdeutsche Elementargrammatik. This version was much more popular and got reprinted at least up to an eleventh edition in 1885. In the preface of the Schulgrammatik, he explicitly cites Becker (see Section 3.2.5) and Billroth (see Section 3.2.7). About Billroth he says: unter den Grammatiken der fremden Sprachen verdanke ich der Lateinischen von Billroth am meisten ‘among the grammars of foreign language I owe most to is the Latin grammar of Billroth’ (Hoffmann 1839: xiii).
[3.90] In the Schulgrammatik Hoffmann includes a graphical analysis of the example sentence in (3.15), as shown in Figure 3.26 (Hoffmann 1839: 148). The graphical display is closely related to Billroth’s display in Figure 3.17. In both graphics, subject and predicate are placed on top, although there is no connecting line in Hoffmann’s version. All other modifiers are linked by lines going down. Note that this example even includes subordinated clauses, namely a relative clause welchen alle bewunderten and an adverbial clause ehe es Mittags geworden war. They are simply treated similar to all other modifiers.
(3.15) | Tapfere Krieger jenes großen Feldherrn, welchen alle bewunderten, erfochten gestern, ehe es Mittags geworden war, eine glänzenden Sieg über ihre zahlreichen Feinde. |
[3.91] Hofmann also uses vertical lines to illustrate the structore of complex sentences consisting of multiple clauses. Interestingly, he separates the discussion of syntactic analysis, using the heading mehrfacher Satz ‘multiple clause’ (1839: 148), from the section on the stilistic analysis, using the heading Periode. However, there does not seem to be a clear separation between the two. The example of a mehrfacher Satz (3.16) is called a Periode. This example is artificially contructed, and in a later edition Hoffmann adds a footnote to excuse for the stilistically poor example:
Die Periode steht hier nur der grammatischen Analyse halber. Stilistisch ist sie mangelhaft, weil viele gleichförmige Satzausgänge unmittelbar auf einander folgen, sobald man die Nebensätze an ihre richtige Stelle bringt. (Hoffmann 1856: 109)
(‘This Periode is only included here for the sake of grammatical analysis. Stilistically she is defective, because many similarly formed subordinate endings follow each other immediately when the subordinate clauses are placed in their rightfull position.’)
(3.16) | Tilly war nicht nur berühmt, weil er, während in anderen Heeren die größte Unorndnung herrscht, die strengste Mannszucht hielt: sondern er konnte auch von sich rühmen, daß seine Soldaten, welche freilich ihren Führer, da dieser so unerbittlich streng war, mehr fürchteten, als liebten, nie eine Schlacht verloren hatte. |
[3.92] Hoffmann uses various devices in the graphical analysis of this complex sentence, though there are many unexplained features and apparently some inconsistencies:
[3.93] The second edition of the Schulgrammatik (Hoffmann 1853) takes over exactly the same graphics, but adds a few more examples of mehrfache Sätze (Hoffmann 1853: 216-217). And then, completely at the end of the book, before the appendices, he adds the following comment in a small typeface, referring to the example sentence “Nr. 5” repeated below in (3.17):
Von Perioden kann man Bilder entwerfen, indem man die Hauptsätze mit A B C, die Nebensätze ersten Grades mit a b c, die Nebensätze zweiten Grades mit α β ɣ, die Nebensätze dritten Grades mit 𝔞 𝔟 𝔠 bezeichnet. So ist z.B. das Schema der eben unter Nr. 5 aufgeführten Periode folgendes: [ A a α A b β : B c. ] (Hoffmann 1853: 240)
(‘It is possible to draft pictures of Perioden by indicating main clauses with A B C, first level subordinate clauses with a b c, second level subordinate clauses with α β ɣ, third level subordinate clauses with 𝔞 𝔟 𝔠. In this way, for example, the scheme of the Periode Nr. 5 is [ A a α A b β : B c. ].’)
(3.17) | Es ist für denjenigen, der mit sich selbst noch nicht recht eins geworden ist, was er denn eigentlich suche und wolle, äußerst beängstigend, wenn er auf seinem Wege auf Sätze stößt, die allen seinen bisherigen Meinungen und den Meinungen seiner Zeitgenossen und der Vorwelt widersprechen: und gewiß ist diese Ängstlichkeit eine der Hauptursachen, warum die Menschheit auf dem Wege zur Wahrheit so langsame Forstschritte gemacht hat. (Hoffmann 1853: 239, citing Fichte) |
[3.94] The use of letters in different typefaces is clearly inspired by Lehmann (see Section 3.2.8), but Hoffmann removes all internal structure of the Periodenbild. The idea to simplify Lehmann’s notation to a single line might have been inspired by Rinne (see Section 3.3.4) or Nägelsbach (see Section 5.1.1). All in all, this Periodenbild feels like an afterthought, which Hoffmann never expanded upon in later editions of the Elementargrammatik.
[3.95] In summary, Hoffmann tried to expand the proposals from Billroth by applying the grammatical line-graphics to subordinate clauses. Although his system is not well documented in his writings, it is comprehensible enough the be understood by experienced linguists. It has inspired Bauer (see Section 3.3.8) and Kern (Section 3.3.24) to further develop the idea of line-based gramatical analyses.
[3.96] Karl Johann Friedrich Rinne (1802-1887) was a teacher in Erfurt and Zeitz. He wrote various textbooks including Die deutsche Grammatik in 1836. This textbook is an introduction to the latest developments in historical comparative grammar written specifically “for teachers and all those, who want to become acquainted with the contemporary position of this science” (für Lehrer und für alle, welche sich mit dem gegenwärte Standpunkte dieser Wissenschaft vertraut machen wollen). However, this textbook does not have any graphical grammatical analyses.
[3.97] Yet, in the first volume of his multi-volume Lehre vom deutschen Stile, published between 1840 and 1847, Rinne uses a system of symbols which he calls Periodensymbolik (Rinne 1840: 276). He explicitly references the “more extensive system” of Lehmann (see Section 3.2.8). Just as Lehmann, Rinne (1840: 272-276) uses capital letters for main clauses and lower-case letters for subordinate clauses. Closer to Herling (see Section 3.2.2), recursively embedded clauses are indicated by superscripts, though using roman numerals. Central to Rinne’s approach is a major separation of each Periode in a Protasis/Vordersatz and an Apodosis/Nachsatz, visually separated by a colon (cf. Diesterweg’s approach in Figure 3.14). Additionally, Rinne uses letters A/a for the protasis and letters B/b for the apodosis. An example is shown in Figure 3.28 for the example sentence in (3.18) from Rinne (1840: 274).
[3.98] The result of Rinne’s graphical choices is a symbolic analysis on a single line, which makes it much easier to include it in a written text compared to the visually more involved approach by Lehmann. The concept of reducing the sentence scheme to a single line is similarly used by various later authors, for example by Nägelsbach (see Section 5.1.1) and by Hoffmann in his second edition (see Section 3.3.3). These similarities could very well have been developed independently, though.
(3.18) | Daß sich alles verändert, was uns hier umgibt; daß sich alles auflöst, was uns hier erscheint; daß wir uns in einem Wirbel befinden, der immer in Bewegung ist, der alles ergreift und mit sich fortreißt, der alles in einem Abgrund stürzt aus welchem es nie wieder hervorkommt: das fühlen wir, so bald wir aufmerken; das fällt in die Augen, so bald wir uns umsehen; das sagt uns unsre Erfahrung, und die Geschichte aller Zeitet bestätigt es |
[3.99] Friedrich Haupt (1805-1891) was a teacher in Zürich and later pastor in the Odenwald region. More biographical information is available in Haupt (1905). During his time in Zürich he wrote various textbooks and prepared an anthology of the German language in his book Deutsche Sprache und Literatur. The second part of this anthology, Deutsche Prosa (Haupt 1841), contains a preface on didactical methods, including syntactic analysis. This preface is removed in later editions of the anthology. Compare the 1865 edition available online at https://mdz-nbn-resolving.de/bsb10597896.
[3.100] In this preface, Haupt is exceedingly enthousiastic about the syntactic approach of Becker. He heralds him as the founder of a new linguistics:
Becker […] Einer der scharfsinnigsten Sprachforscher, der der Begründer eines neuen Sprachsystems, und einer wahrhaft formal bildenden Unterrichtsmethode geworden ist (Haupt 1841: 295)
(‘Becker […] One of the most astute linguists, who has become the founder of a new language system and a true formally-educating method of teaching.’)
[3.101] In a footnote Haupt even announces that he will publish an extensive analysis of Becker’s method in a book to be called K. F. Becker und die deutsche Volksschule. Ein Beitrag zur Förderung deutscher Sprachwissenschaft und Methodik (Haupt 1841: viii). Unfortunately, he does not appear to have ever finished this planned book.
[3.102] As an example of Becker’s syntactic analysis, Haupt analysed the example sentence (3.19) from Jean Paul. In a first step he separates the various clauses and illustrates their interrelation with the graphical display in Figure 3.29. Subsequently, Haupt also analyses the internal structure of one of the main clauses, but this is only done in words. Note that Haupt’s graphical approach, using hierarchically-ordered reversed horizontal braces, is not found in the work of Becker (see Section 3.2.5). However, it is reminiscent of the visual approach by Diesterweg in Figure 3.16.
(3.19) | Der wahnsinnige Sterbliche begeht seine Sünden so kühn, blos weil sich ihm ihre mörderischen Folgen verhüllen; er kettet die in seiner Brust eingesperrten reißenden Thiere los und lässet sie in der Nacht unter die Menschn dringen; aber er siehet es nicht, wie viele Unschuldige das losgebundene Unthier ergreife und würge. (Haupt 1841: viii, citing Jean Paul) |
[3.103] The abbreviations that are used in the graphical analysis can be reconstructed from the text as follows:
[3.104] Karl Wilhelm Eduard Mager (1810-1858) was a school teacher and educator. He was a close collaborator of Diesterweg (see Section 3.2.6). Among Mager’s many works there is also a textbook about German grammar, called Deutsches Sprachbuch. Anfänge der Grammatik, Onomatik und Sprachkunst (Mager 1842). In this book he uses a quite elaborate system of graphical syntactic analysis for complex sentences, illustrating his system with many examples (Mager 1842: 115-119). Mager uses multiple visual clues to illustrate the structure of a sentence, many of which appear to be inspired by Hoffmann (see Section 3.3.3), although Hoffmann is not cited by Mager. Mager’s analytic system is illustrated here in Figure 3.30 for the multi-clause sentence (3.20).
[3.105] First, each clause is placed on a separate line. Second, clausal subordination is indicated by indentation, with multiple subordination leading to multiple indentation. Additionally, sometimes vertical lines are added for subordination, but it is not explained when to do this – and the examples given are also inconclusive. Third, he uses arbitrary letters for different kinds of subordination (a=Attributivsatz, b=Substantivsatz, c=Adverbialsatz). Fourth, secondary subordination is indicated with lower-case greek letters αβγ and third-grade subordination is marked by lower-case Fraktur letters 𝔞𝔟𝔠. Fifth, non-finite clauses are indicated by an additional apostrophe. Finally, coordinated clauses have an additional numeral.
(3.20) | Sein Vater, a1 der ihn herzlich liebte, und a2 Alles aufbot γ’ um ihm eine gute Erziehung zu geben, aber doch a3 wenig Freude an ihm erlebte, mußte sich endlich entschließen, b’1 ihn zu einem strengen Lehrherrn zu thun und b’2 diesen zu bitten, β’ den Taugenichts in harter Zucht zu halten. |
[3.106] Before the German textbook, Mager already wrote a French textbook, originally published 1840, but cited here from fourth edition (Mager 1847: 316-322). Exactly the same edition is reprinted at least up to the seventh edition (Mager 1854: 316-322). Exactly the same system that was used for German multi-clause sentences is used here for French sentences, like (3.21 a) analysed in Figure 3.31. This example is a quotation from the Sermon pour le premier dimanche de carême of Jean-Baptiste Massillon. However, the precise formulation of the quotation indicates that this example is not quoted directly from Massillon, but indirectly from the Cours d’études of Étienne Bonnot de Condillac. Remarkably, Condillac cites this example as une période qui n’est pas si bein faite, parce qu’il y a trop de propositions incidentes dans le premier membre ‘a period that is not so well constructed, because there are too many auxiliary propositions in the first member’ (Condillac 1775: 87-88).
(3.21) | a. | Souvenez-vous de ce jeune roi de Juda a1 qui, γ’ pour avoir préféré les avis d’une jeunesse inconsidérée à la sagesse et à la maturité de ceux 𝔞1 aux conseils desquels Salomon son père étoit redevable de la gloire et de la prospérité de son règne, 𝔞2 et qui lui conseilloient d’affermir les commencements du sien par le soulagement des ses peuples, a1 cont. vit un nouveau royaume se former des débris de celui de Juda a2 et qui, γ pour avoir voulu exiger de ses sujets au delà de 𝔟 ce qu’ils lui devaient, a2 cont. perdit leur amour et leur fidélité α qui lui étaient dûs. |
b. | Remember that young king of Judah a1 who, γ’ for having preferred the advice of a thoughtless youth above the wisdom and maturity of those 𝔞1 to whose counsel Solomon, his father, was indebted for the glory and prosperity of his reign, 𝔞2 and who advised him to strengthen his own start by the relief of his people, a1 cont. saw a new kingdom formed from the ruins of that of Judah a2 and who, γ for having wanted to demand from his subjects beyond 𝔟 what they owed him, a2 cont. lost their love and their fidelity α which were due to him. |
[3.107] Johann Baptist Fürg (1802-1864) was a teacher in München. Not much is known about him, except for a glowing necrolog (Heitz 1864). He wrote various small booklets for use in school, among them a four volume grammar Die deutsche Sprache dargestellt in Aufgaben sammt den nöthigsten Erläuterungen für die Volksschulen Bayerns (Fürg 1850). However, in that work there are no graphical methods used.
[3.108] Earlier he wrote smaller textbook Die Satz- und Styllehre mit Inbegriff der Wortbiegung, dargestellt in 760 Aufgaben sammt den nöthigsten Erläuterungen, which was first published in 1845. In the fourth edition of 1847 he expanded the preface (cf. the third edition without it, Fürg 1846). In this new preface, he addresses the teachers and presents to them a few examples of how to analyse sentences (Fürg 1847: vi-viii). One of these examples is shown in Figure 3.32. This numbered tabular analysis is strongly reminiscent of Becker’s approach (see Section 3.2.5). Becker is not mentioned in Fürg’s text, but Fürg calls this approach Analyse or Zergliederung, and these are both words used by Becker. It is also telling that these examples are directed to teachers, not students, which is again similar to Becker. Finally, Fürg calls himself an adherent of the “new grammar”, which might be a reference to Becker’s approach to grammar:
Aus dem bisher Gesagten geht zur Genüge hervor, daß ich mich entschieden als Anhänger der neuen Grammatik bekenne, und daß ich sie als den einzigen Weg betrachte, der zum wahren Verstehen der Sprache führt. (Fürg 1847: viii)
(‘It is clear from what has been said upto now, that I resolutely declare myself to be an adherent of the new grammar, and that I consider it to be the only way that will lead to a real understanding of language.’)
[3.109] Friedrich Bauer (1812-1874) was the founding director of a missionary education centre in the Bavarian city of Neuendettelsau. This institute is still active today as Mission EineWelt. Among other duties, Bauer also taught religion and German grammar at his institute (Fuchshuber-Weiß 2013: 93). In that context he wrote the Grundzüge der neuhochdeutschen Grammatik für die unteren und mittleren Klassen höherer Bildungsanstalten, which was first published in 1850. Bauer himself delivered various editions of this book up to his death in 1874.
[3.110] After Bauer’s death, the German dictionary maker Konrad Duden kept producing new editions of this grammar, with only very minor changes. The Neuhochdeutsche Grammatik remained popular and has seen at least 27 editions, including special editions for Austria, for protestant schools, and for catholic schools. The last edition was published in 1912 (see Fuchshuber-Weiß 2013: 115-117, 128-131 for a survey of all editions). However, even that was not the end: Bauer’s text was given a second life in the context of the expanding Duden “empire”. In the century since the death of Konrad Duden in 1911, the name Duden has become a household name in Germany for everything related to the German language. The 11th edition of the Duden dictionary in 1935 was expanded with a German grammar, prepared under the auspice of Otto Basler (Basler 1935). This grammar was basically a new edition of Bauer’s Neuhochdeutsche Grammatik (without clear attribution). Only after the Second World War the Duden grammar was completely rewritten. Updated editions of this rewritten grammar are still popular today in Germany.
[3.111] In the grammatical analysis of complex sentences, Bauer (1850: 94-97) uses a graphical display that is eerily similar to the one proposed by Hoffmann (see Section 3.3.3). Although Hoffmann is not cited by Bauer, it is known that Bauer had a copy of Hoffmann’s grammar in his personal library (Fuchshuber-Weiß 2013: 99), so an influence seems highly likely. The first few examples of Bauer’s syntactic analysis are shown in Figure 3.33 (Bauer 1850: 95). These examples are also found in exactly the same layout 85 years later in Basler (1935: 254). Each sentence is analyzed in two ways: the text itself is subdivided into clauses, with subordinate clauses shifted down vertically and connected with lines, very similar to Hoffmann (see Figure 3.27).
[3.112] Additionally, Bauer adds structural analyses in the form of letters, with capital letters for main clauses and lower-case letters for subordinate clauses. First-level subordination is separated by a comma, while secondary or deeper subordination is indicated by vertical lines and numerical superscripts. Bauer uses different lower-case letters for different kind of subordinate clauses (Bauer 1850: 94), though the assignemt of the letters is completely unmnemonical, namely a for complement clauses (Substantivsatz), b for relative clauses (Adjectivsatz) and c for adverbial clauses (Adverbialsatz). The so-called verkürzte subordinate clauses (i.e. zu+Infinitiv constructions) have dashes -a-, -b-, -c-, but Bauer explains in a later edition that this was supposed to be a strikethrough (Bauer 1863: footnote on p. 145).
[3.113] A more complex example discussed by Bauer is the sentence in (3.22), analyzed as Figure 3.34 (Bauer 1850: 97). Again, there is both an analysis of all main and subordinate clauses, written out with subordination marked by vertical lines, and there is a structural pattern with letters and symbols in the bottom right. This structural pattern has two cases of coordination (marked by a plus symbol), an internal subordination (marked by the brackets), a secondary subordination (marked by a vertical line) and a subordination at the end of the sentence (marked by a comma). Probably it is more accurate to analyze the subordination c as a simple frontal embedding of the second conjunct, using a comma in the analysis instead of brackets for a clause-internal embedding (viz. A+c,A,a+a instead of A+A(c)A,a+a), but that is some minor nitpicking.
(3.22) | Er schenkte ihr 300 Rubel und als er erfuhr, daß sie in Petersburg einen Vetter habe, stellte er ihre frei, ob sie ihre Reise fortsetzen oder ob sie umkehren wolle. (Bauer 1850: 97, citing an example from Johan Peter Hebel Herr Charles) |
[3.114] At least since the 6th edition, Bauer expands the notation of the symbolic formulas with various additional elements (Bauer 1863: 144-150):
[3.115] In this edition he also promises to write a more extensive explanation of his notation, but he never seems to have done this:
Die Darstellung solcher Satzbilder durch Formeln ist äußert lehrreich und förderlich, sowohl für das Verständnis schwerer Sätze, als für das Schreiben. […] Solche Formeln zur Darstellung von Satzbildern zu geben, muß sich der Verfaßer vorbehalten für eine besondere kleine Schrift, welche eine Ergänzung der Syntax für eine höhere Lehrstufe bilden soll. (Bauer 1863: 128)
The presentation of such Satzbilder through formulas is very instructive and beneficial, both for the understanding of difficult sentences and for the writing. […] The demonstration of such formulas of the Satzbilder will be reserved for a special small booklet, which will be an extension of the syntax for the for the higher classes.
[3.116] Karl Ferdinand Hermes (born 1820) was director of a privately run school in Berlin, called Höhere Töchterschule zum heiligen Kreuz. Annual school-report of 1877: https://google.com/books?id=oS4-AQAAMAAJ. He was also Chairman of the Verein für höhere Töchterschulen and Editor-in-Chief of the journal Stoa: Zeitschrift für die Interessen der höheren Töchterschulen. Digital versions of the journal Stoa from 1868 and 1869: https://scripta.bbf.dipf.de/viewer/toc/020698577. A short biographical profile is given by Schütze (2014: 940).
[3.117] He wrote a German grammar with the name Unsere Muttersprache, which was first published in 1852, though it will be quoted here from the fourth edition of 1866 (Hermes 1866). The book reappeared at least until the tenth edition of 1881. He discusses the technique to use a Satzbild for sentence analysis in the section der zusammengesetzte Satz (Hermes 1866: 102, 105). The details of the notation are rather idiosyncratic, though it is clearly inspired by the concept of the Periodenbild of Lehmann and Götzinger. Knowledge about this approach appears to be rather widespread among German teachers by this time, and it is often difficult to reconstruct exactly who has been their direct inspiration.
[3.118] Hermes analyses the sentence in (3.23) with the graphic shown in Figure 3.35. He uses a capital H for main clauses (Hauptsatz) and the lower-case letters s for complement clauses (substantivisch), r for relative clauses and a for adverbial clauses. Inserted clauses (Zwischensätze) are indicated by round brackets and secondary subordination is depicated by shifting the letter down. There are no examples with more complex multiple embeddings.
(3.23) | Theophanes zerstörte mit seinem griechischen Feuer die russische Flotte, welche Igor in der sichern Hoffnung, dass der Sieg sein werde, mit großem Aufwande nach Constantinopel geführt; doch kehrte derselbe, ehe zwölf Jahre vorüber waren, um so furchtbarer wieder. (Hermes 1866: 105) |
[3.119] Karl Friedrich Wilhelm Wander (1803-1879) was a teacher in Silesia and a progressive padagogue in the tradition of Pestalozzi. He was persecuted by the prussian government as a revolutionist and communist, though his views appear moderately liberal-progressive from a contemporary point of view. He was a close friend of Diesterweg (see Section 3.2.6) and wrote an influential pamphlet in defence of his friend, called Der geschmähte Diesterweg. In German linguistics, he is most well-known for his collection of German proverbs, the Deutsches Sprichwörter-Lexikon. Extensive biographical information about Wander is provided by Hohendorf (1979).
[3.120] Among Wander’s many publications there is a small booklet called Satzbilderschule (Wander 1856). It does not seem to have been noted by his contemporaries, and today only very few copies remain available. Wander’s analysis of Satzbilder closely follows Diesterweg in all its details (see Section 3.2.6). Wander’s book is organised by possible sentence structures, trying to enumerate all possibilities of complex sentences. With this approach, Wander follows the same plan as Lehmann (see Section 3.2.8). Both Diesterweg and Lehmann are explicitly mentioned in the preface (Wander 1856: vi).
[3.121] The main part of Wander’s book consists only of Satzbilder in the form of letter sequences. Each letter encodes a clause, capital letters for main clauses and lower-case letters for subordinate clauses. Superscript numbers indicate the level of embedding. However, in the preface Wander additionally discusses three different kinds of visualisation that can be used to enhance the bare-bones letter-based analysis. The first two visualisations are directly adapted from Diesterweg, but the third appears to be an innovation of Wander himself.
[3.122] For the analysis of the example sentence in (3.24), Wander proposes two visualisations. First, in Figure 3.36, he combines the rising-falling analysis of the Kommata in a Periode with the letter sequences of the individual clauses. Second, in Figure 3.37 he uses a grid to show the sequence of clauses from left to right and the depth of embedding from top to bottom. Another example of both these visualisations is found at the end of his book (Wander 1856: 63-64).
[3.123] For the analysis of the example sentence in (3.25) Wander uses another visualisation, shown in Figure 3.38. He displays the hierarchical structure by indentation and connects separated parts of the same clause by large brackets. In all these example-visualisations there are minor typographic errors, as discussed below in the captions to the figures. With such complex sentences it turns out to be really difficult to stay faultless, even for an obviously competent author like Wander.
(3.24) | a Wer nur den groß nennt, a2 der in einem ungewöhnlichen Grade alles ist, a3 was er soll; b wer aus der Zahl großer Monarchen jeden ausstößt, b2 dessen Regierung nicht durch ihn selbst, c2 sondern nur durch das glückliche, einträchtige Genie vortrefflicher Diener glänzte und d2 der nur weise genug war, b3 sich leiten zu lassen, a4 da er selbst hätte leiten sollen; c wer, e2 mit unverwandtem Blick auf den einzigen würdigen Zweck eines Königs, c keine, auch nicht die glänzendsten Thaten bewundert, f sobald sie jenem Zweck entgegenlaufen; d wer das einseitige Talent eines Kriegers von dem mannichfaltigen, so viele andere Talente in sich schließenden eines Monarchen unterscheidet: A der wird die großen Könige, e groß im echten Sinne des Wortes, A durch ganze Jahrhunderte und unter ganzen Nationen vergebens suchen; B er wird schon, a eh’ er sucht, B ihrer nur äußerst wenige zu finden hoffen. (Wander 1856: vii-viii, citing Johann Jakob Engel) |
(3.25) | (A) Die lebhaften Schilderungen (a) welche die Tarentiner ihm von dem glücklichen Leben machten, (a2) das im ruhigen Schoße ihres Vaterlandes und in Gesellschaft seiner dortigen Freunde auf ihn warte, (A) vollendeten endlich die Wirkung, (b) die der gewaltsame Zustand, (b2) worin er seit einiger Zeit gelebt hatte, (b) auf ein Gemüth, (c2) wie das seinige, (b) machen mußte, (c) indem sie ihm zugleich den ganzen Widerwillen, (d2) den er nach seiner Verbannung von Athen gegen den Stand eines Staatsmannes gefaßt hatte, (c) und seinen ganzen Hang zur Abgeschiedenheit von der Welt und zum Leben mit sich selbst und mit guten Menschen wieder gaben, (e2) welches ihm, (a3) wie er glaubte, (e2) jetzt um so nöthiger war, (b3) da er sein Gemüth auch von den geringsten Rostflecken, (a4) die von jenem syrakusischen Hofleben zurückgeblieben sein könnte, (b3) zu reinigen wünschte. (Wander 1856: x, citing C. M. Wieland Geschichte des Agathon. Clause identifiers were not included in the original by Wander, but were added here for easier interpretation) |
[3.124] Heinrich August Schötensack (1812-1891) was a teacher in Stendal. He wrote a massive (more than 800 pages) Grammatik der neuhochdeutschen Sprache. It contains just a very short discussion of Periodenbilder without much detail (Schötensack 1856: 800-802). The example in Figure 3.39 for the sentence in (3.26) clearly shows the influence from Lehmann (see Section 3.2.8) in the use of the different typefaces for the different levels of embedding. However, even Lehmann did not venture into the Hebrew characters, as is shown here by the use of the Alef for the fifth embedding.
(3.26) | Er gefiel sich sehr in Scherzen, (a) weil er wusste, (α) dass Nichts mehr den Geist auſheitre, (β) Nichts mehr eine freimüthige Aeusserung gestatte, (𝔄) als der Scherz, (𝔞) der fast immer gut aufgenommen werde, (א) wenn nicht gerade ein murrköpfiger Thor ihn entgegen zu nehmen habe. (Schötensack 1856: 801) |
[3.125] Raymund Schlecht (1811-1891) was director of a teacher’s training college in Eichstätt (see Gmelch 1931 for biographical details). Besides being a teacher and theologian, he was a musical researcher. Most of his writings deal with musicology, but he also published a short grammar called Denk- und Sprachlehre. Ein Leitfaden zur Ertheilung eines organischen Sprachunterrichts (Schlecht 1856). In the preface he prominently mentions Becker (see Section 3.2.5).
[3.126] The grammar contains a few graphical displays for complex sentences using a musical notation (Schlecht 1856: 71, 79). This is reminiscent of the approach by Herling, as shown in Figure 3.8, and Diesterweg, as shown in Figure 3.15. There is no indication that Herling was an inspiration for Schlecht. In contrast, Schlecht clearly knew about Diesterweg, whom he went to visit in Berlin in 1862 (according to Gmelch 1931: 21) and about whose didactic methods he wrote a small note in the Katholische Schulzeitung in 1868 (Gmelch 1931: 54). The small note appeared in two parts, which are available online at https://www.digitale-sammlungen.de/view/bsb10687402?page=260 and https://www.digitale-sammlungen.de/view/bsb10687402?page=284. Yet, given that Schlecht was more of a musicologist than a linguist, it seems just as likely that he independently came up with the idea of using the musical score for the depiction of linguistic structure.
[3.127] He uses the musical display in Figure 3.40 to illustrate the structure of the highly complex sentence in (3.27). Before the main clause there are eleven subordinated clauses, separated graphically by a double bar. These eleven subordinated clauses are again subdivided into three subsections devided by a single bar, each of which is again internally complex. The main clause, after the double bar, also has a complex internal structure. It is impressive that grammarians like Schlecht had the audacity to tackle such complex sentences, something rarely seen in contemporary linguistics.
(3.27) | Wie das innere Bewegen, das im Herbst die Schaaren der wandernden Vögel mit unwiderstehlicher Gewalt ergreift und über Land und Meer dahinführt, wie der Trieb der Säfte, der in den Tagen des Frühlings die Tausende der Blüthen aus dem Fruchtbaum herausdrängt an’s Licht, an welchem die meisten, gleich einer vergeblichen Ueberfülle, verblühen und abfallen, ohne Frucht zu tragen, wie jeder Zug des Instinkts im niederen Reiche der Sichtbarkeit ein Walten der Kraft ist, die das Wesen der Leiblichkiet begründete und beherrscht; – so war auch jenes Bewegen, das die Menschenseelen zu dem Werke der Kreuzzüge dahin riß, nicht von unten her, sondern es kam von oben, aus einem Anwehen des Geistes, der dem Menschen im Anfang den lebendigen Odem gab. (Schlecht 1856: 78, citing Gottfried Heinrich Schubert Reise in das Morgenland) |
[3.128] Ferdinand Bachmann (1817-1891) was a teacher and later director of a teachers’ seminary in Prague (more biographical details in Jahne 1897). He was chosen for this job partly because of the popularity of his lectures, published 1864 as a textbook Die Satz- und Wortlehre in praktischen Vorträgen. Ein methodischer Leitfaden zur Behandlung des grammatischen Theils unserer Lesebücher in der Volksschule. Für angehende Lehrer und Lehrerinnen (Bachmann 1868, cited here from the second edition). The book saw reprints at least up to a fifth edition in 1884.
[3.129] Bachmann uses Satzbilder to illustrate the structure of complex sentences, copying the style from Bauer (see Section 3.3.8), who is not cited. The construction of this graphical display is explained in great detail (Bachmann 1868: 97-104) and exemplary applied for various longer sentences in an appendix (Bachmann 1868: 244-255). An example is given below in Figure 3.41 for the example sentence in (3.28). The graphical analysis only deals with the clause structure, but note that conjoined predicates like schön und edel ‘beautiful and noble’ are treated as two conjoined clauses marked with a plus-symbol and a large arch above. Commas are used both for preposed and postposed subordination, while internal subordination is marked by round brackets (not shown in the figure). Secondary and deeper subordination is placed downwards connected by lines and additionally marked by superscript numbers. Strikethrough indicates non-finite clauses.
(3.28) | a+a, Wenn es schön und edel ist, s̶2 für das Vaterland zu sterben; A+A so ist es noch schöner und edler, ,s̶ so ganz für das Vaterland zu leben. (Bachmann 1868: 250, with obvious reference to Horace’s dulce et decorum est pro patria mori) |
[3.130] Bachmann also uses tables, with the each word as a row and two columns for the analysis: one row is called syntaktisch (for the syntactic function) and the other row is called etymologisch (for the morphological analysis). This approach, including the term “etymology” for what we would today call morphology, is highly reminiscent of the british tradition of parsing as proposed by Murray (1824: 195). It could also be an influence from Becker (see Section 3.2.5), but Bachmann does not use numbers to explicate the dependencies. This table-style approach is first introduced only for syntactic functions (Bachmann 1868: 38), but later applied in great detail in the appendix including morphological analysis (Bachmann 1868: 244-255). A tabular analysis of the first subordinate part of example (3.28) is showhn in Figure 3.42.
[3.131] Eduard and Friedrich Wetzel were brothers and both teachers in Berlin. Friedrich Ludwig Eduard Wetzel (1819-1895) was a teacher at the Lehrerinnenseminar and the Augustaschule and Friedrich Wilhelm Eduard Wetzel (1824-1897) was teacher at the Elisabethschule. A few more biographical details are available in the Personenregister of Schütze (2014: 885-886).
[3.132] In 1865 they wrote a textbook on German grammar, called Die deutsche Sprache. Eine nach methodischen Grundsätzen bearbeitete Grammatik für höhere Lehranstalten und zum Selbstunterricht, which was reprinted at least until the 13th edition of 1914. This book is cited here in the third edition (Wetzel & Wetzel 1871), which is the earliest version that I have been able to access. From the prefaces of the earlier version it does not appear that the graphical analysis has seen much change since the first edition. The same analysis is also found in the first edition of the shortened version, called Leitfaden für den Unterricht in der deutschen Sprache (Wetzel & Wetzel 1868: 182). This shorter book has also been reprinted numerous times, retaining the same graphical analyses, at least up to a 46th edition (Wetzel & Wetzel 1893: 223-224).
[3.133] Their analysis of multi-clause sentences is limited to inserting symbols into the text at the end of each clause (Wetzel & Wetzel 1871: 313-314), similar to Herling (1823) as shown in Figure 3.6. Wetzel & Wetzel use capital letters for main clause and lower-case letters for subordinate clauses, with superscript numbers for the depth of the embedding. All subordinate clauses of a specific main clause use the same letter, e.g. all A/a or B/b etc. An example of their analysis is shown in Figure 3.43 for the example sentence in (3.29).
(3.29) | Es ist ein altes Sprichwort (A), daß der Mensch dann beten lernt (a1), wenn er in eine Noth geräth (a2), aus der er sich selbst nicht zu erretten vermag (a3), weil es ihm an der rechten Einsicht fehlt (a4), die für diesen Zweck ersprießlichen Mittel und Wege zu wählen (a5). |
[3.134] Josef Lehmann (1838-1911) was a teacher at the teachers’ seminary in Vienna (Austria), not related to Johann Lehmann (see Section 3.2.8). A few biographical details about Josef Lehmann are available in Klein (1970). Lehmann wrote various popular textbooks that all saw many different editions:
[3.135] All visual methods for grammatical analysis used by Lehmann are already present in the Leitfaden, in which he cites Bauer (see Section 3.3.8) and Götzinger (see Section 3.3.1) among his sources (Lehmann 1870: v). Lehmann uses three different approaches for syntactic analysis. First, he uses a tabular analysis like Bachmann (see Section 3.3.13), which is called Wort- und Satzanalyse. Shown in Figure 3.44 is only the first part of a long sentence analysed by Lehmann (1870: 186-187). As discussed for Bachmann (cf. Figure 3.42), this might be an influence from Becker (cf. Figure 3.11), though crucially the idea of using numbered rows is missing.
[3.136] Second, for complex sentences Lehmann uses a letter-based Satzbild (Lehmann 1870: 217-218; see also Schulgrammatik: Lehmann 1878a: 281-282; and Sprachbuch: Lehmann 1878b: 128). He use a simplified one-line approach with lower-case letters indicating the function of the subordinate clause (e.g. o=Objektsatz, k=Kausalsatz, m=Modalsatz). Depth of embedding is indicated by a number after the letter and non-finite clauses by an additional letter v=verkürzt. An example is shown in Figure 3.45 for the example sentence in (3.30).
(3.30) | A Thu’, o was jeder loben müsste, k2 wenn die ganze Welt es wüsste; B thu’ es, m dass es niemand weiß, und C gedoppelt ist dein Preis. (Lehmann 1870: 218, citing an aphorism of Friedrich Rückert) |
[3.137] Finally, for even larger syntactic complexes Lehmann uses a Bild der Periode (Lehmann 1870: 223-224; see also Schulgrammatik: Lehmann 1878a: 288). This graphical display, as shown in Figure 3.46 for the Periode in (3.31), is obviously similar to the approach of Diesterweg (see Section 3.2.6). However, it might also have been an influence from Wander (see Section 3.3.10), who himself was influenced by Diesterweg. The separation of the Periode in a rising and a falling part is indicated by the slanted lines, the letters represent the various clauses.
(3.31) | s Dass die goldenen Tage der harmlosen Kindheit so schnell verschwinden; s dass die Blüten unsrer edelsten Freuden so bald abfallen; s dass Freunde, a die es oft am treuesten meinen, s am ehesten von unsrer Seite geriffen werden: A das versetzt den gefühlvollen Menschen nicht selten in stille Wehmuth. |
[3.138] Joseph Zitzlsperger (1853?-1893) was a teacher in Amberg. He wrote textbooks about accounting, history and German language. His book Leitfaden für den Unterricht in den Anfängen der deutschen Sprachwissenschaft an Mittelschulen (Zitzlsperger 1871: 133-135) contains is a short discusion about Satzbilder, very much in the tradition of the Periodenbilder of Bauer (see Section 3.3.8).
[3.139] The use of the symbols clearly shows a direct influence from Bauer, especially the variant from Bauer’s 1863 revision, although this influence is not explicitly mentioned by Zitzlsperger. The main aspect of the graphical display that is added by Zitzlsperger is the underlining of the letters. A single underline indicates a bekleidete ‘dressed’ clause, while a double underline stands for a zusammengezogene ‘drawn together’ clause (roughly a clause that is the result of conjunction reduction, i.e. a clause-internal conjunction that could be analysed as the result of two clauses with duplicated elements removed). Basic subject+predicate clauses, i.e. nackte ‘naked’ clauses, are not underlined
[3.140] Another difference in Zitzlsperger’s notation is the meaning of superscript numbers. Bauer used them to indicate the depth of embedding (in the tradition of Herling, see Section 3.2.2). However, the depth of the embedding is also indicated by the vertical displacement, i.e. graphically lower placement of the letter indicates the depth of the embedding. Omitting the double marking of depth, Zitzlsperger repurposed the superscript numbers to refer to different kinds of adverbial clauses. However, the meaning of the numbers are now completely arbitrary, e.g. 1 stands for locative clauses and 2 for temporal clauses, 3 for comparative clauses, etc. (Zitzlsperger 1871: 140).
(3.32) | Es ist ein Lob für einen Mann, wenn man seine Fehler sagen darf, ohne daß er aufhört, groß zu sein. (Zitzlsperger 1871: 135) |
[3.141] Adolf Schütze (born 1824) was a teacher at the Karls-Gymnasium in Bernburg. In the school’s yearbook of 1875 he wrote a small essay Über den deutschen Gliedersatz (Periode) und die Gesetze seines Baues (Schütze 1875), which includes an interesting graphical syntactic analysis. He analyses a long quote from Friedrich Schiller (3.33), consisting of four sentences with numerous subordinate clauses. This long and complex linguistic structure is very concisely summarised by the graphic in Figure 3.48. In a footnote he cites Herling, Lehmann, Becker and Haupt (Schütze 1875: 8), but the graphical display is more similar to Diesterwegs’s approach (see Figure 3.15). However, the details are completely idiosyncratic:
(3.33) | a. | Möglich allerdings und sehr wahrscheinlich, dass die französischen Protestanten emsig daran arbeiteten, in den Niederlanden eine Pflanzschule für ihre Religion zu erhalten, und eine gütliche Vergleichung ihrer dortigen Glaubensbrüder mit dem Könige von Spanien durch jedes Mittel zu verhindern strebten, um diesem unversöhnlichen Feinde ihrer Partei in seinem eigenen Lande zu thun zu geben; |
b. | sehr natürlich also, dass ihre Unterhändler in den Provinzen nicht unterlassen haben werden, die unterdrückten Religionsverwandten zu verwegenen Hoffnungen zu ermuntern, ihre Erbitterung gegen die herrschende Kirche auf alle Arten zu nähren, den Druck, worunter sie seufzte, zu übertreiben, und sie dadurch unvermerkt zu Unthaten fortzureissen. | |
c. | Möglich, dass es auch unter den Verbundenen Viele gab, die ihrer eigenen verlorenen Sache dadurch aufzuhelfen meinten, wenn sie die Zahl ihrer Mitschuldigen vermehrten; die die Rechtmässigkeit ihres Bundes nicht anders retten zu können glaubten, als wenn sie die unglücklichen Folgen wirklich herbeiriefen, wovor sie den König gewarnt hatten, und die in dem allgemeinen Verbrechen ihr eigenes zu verhüllen hofften. | |
d. | Dass aber die Bilderstürmerei die Frucht eines überlegten Planes gewesen, der auf dem Convente zu St. Truyen verabredet worden; | |
e. | dass in einer solennen Versammlung so vieler Edlen und Tapfern, unter denen noch bei weitem der grössere Theil dem Papstthum anhing, ein Rasender sich hätte erdreisten sollen, den Entwurf zu einer offenbaren Schandthat zu geben, die nicht sowohl eine abgesonderte Religionspartei kränkte, als vielmehr alle Achtung für Religion überhaupt und alle Sittlichkeit mit Füssen trat, und die nur in dem schlammigen Schosse einer verworfenen Pöbelseele empfangen werden konnte: | |
f. | (dies Alles) wäre allein schon darum nicht glaublich, weil diese wüthende That in ihrer Entstehung zu rasch, in ihrer Ausführung zu leidenschaftlich, zu ungeheuer erscheint, um nicht die Geburt des Augenblicks zu sein, in welchem sie ans Licht trat, und weil sie aus dem Umständen, die ihr vorhergingen, so natürlich gliesst, dass es so tiefer Nachsuchungen nicht bedarf, um ihre Entstehung zu erklären. (Schütze 1875: 14-15, citing Schiller Geschichte des Abfalls der Niederlande) |
[3.142] Theodor Gelbe (1837-1892) was a school director in Stollberg. He wrote a textbook about German grammar called Deutsche Sprachlehre für höhere Lehranstalten sowie zum Selbststudium (Gelbe 1877). Apparently this book was not very popular, as it does not seem to have been reprinted. A reviewer was very critical of the principle of using a graphical display for the analysis of sentences, which prompted Gelbe to write a quite ironical rebuttal (Gelbe 1880). In this paper, Gelbe lists various other examples of Satzbilder (citing Lehmann, Götzinger, Hoffmann, Bauer and Wetzel), arguing that this approach is widespread and used by respected scholars. It is clear from his writing, that the idea of Satzbilder or Periodenbilder is well-known in the community of grammatical researchers by this point in the 19th Century, and that each scholar, including Gelbe himself, uses slightly different conventions:
In der Entwerfung der Satzbilder befolgt man verschiedene Grundsätze, deren keiner jedoch bis jetzt sich allgemeine Anerkennung zu verschaffen vermochte. (Gelbe 1877: 150)
(‘Different principles are used in the design of Satzbilder, none of which has been able to obtain widespread recognition.’)
[3.143] Gelbe’s approach (1877: 150-153) seems to be most like the approach from Götzinger (see Section 3.3.1). Like Götzinger, Gelbe uses different letters for each clause. However, clause-final embedding is indicated by a straight line with the embedding placed below, while Götzinger uses a horizontal brace. Galbe uses the horizontal brace for various kinds of coordination. The superscript v stands for Verkürzung, i.e. shortened. As an example, Gelbe’s analysis of the example from Lessing in (3.34) is given here in Figure 3.49.
(3.34) | Wenn uns Gott durch einen seiner Engel – ist zu sagen, durch einen Diener seines Worts – ein Mittel bekannt zu machen würdiget, das Wol der ganzen Christenheit, das Heil der Kirche auf eine ganz besondre Weise zu fördern, zu befestigen: wer darf sich da noch unterstehn, die Willkür des, der die Vernunft erschaffen, nach Vernunft zu untersuchen? und das ewige Gesetz der Herrlichkeit des Himmels nach den kleinen Regeln einer eiteln Ehre zu prüfen? (Gelbe 1877: 148, citing Lessing’s Nathan) |
[3.144] Joseph Diringer (1878-1883) was a teacher for German, Latin and Greek in Eichstätt. Besides various editions of classical texts, he wrote Die Periode oder der Gliedersatz in der deutschen Sprache (Diringer 1878). The second half of this small booklet is devoted to the application of Periodenbilder to the analysis of literary sentences.
[3.145] At the end of the preface to his book, Diringer cites many of the graphical approaches that are discussed above (e.g. Bauer, Becker, Götzinger, Herling, Lehmann), but the direct influences on Diringer’s graphical display seem to be Schlecht (see Section 3.3.12) and Schütze (see Section 3.3.17). He uses a quasi-music-score approach with horizontal lines for the depth of embedding and vertical lines for the separation of main sentences. Different lower-case letters are used for distinct kind of subordination, e.g. s for Subjektivsatz ‘subject complement clause’ and l for Lokalsatz ‘local clause’, etc. As an example of his approach, the sentence from Schiller in (3.35) is analysed by Diringer as shown in Figure 3.50.
(3.35) | Wie der bildende Künstler die faltige Fülle der Gewänder um seine Figuren breitet, um die Räume seines Bildes reich und anmutig auszufüllen, um die getrennten Partieen desselben in ruhigen Massen stetig zu verbinden, um der Farbe, die das Auge reizt und erquickt, einen Spielraum zu geben, um die menschlichen Formen zugleich geistreich zu verhüllen und sichtbar zu machen: ebenso durchflicht und umgibt der tragische Dichter seine streng abgemessene Handlung und die festen Umrisse seiner handelnden Figuren mit einem lyirschen Prachgewebe, in welchem sich, als wie in einem weitgefalteten Purpurgewand, die handelnden Personen frei und edel mit einer gehaltenen Würde und hoher Ruhe bewegen. (Diringer 1878: 48, citing Schiller’s Über den Gebrauch des Chors in der Tragödie) |
[3.146] Franz Esser (dates unknown) was a teacher in Weissenburg. In the yearly report of his Gymnasium for the year 1877-1878 he writes a short article with the title Abhandlung über die Formen der Periode im Nibelungenlied (Esser 1878). In this article he analyses sentences from the Nibelungenlied by using Periodenbilder. He organises the article by the complexity of the sentences.
[3.147] For the inspiration of the Periodenbilder he cites Götzinger (see Section 3.3.1). However, different from Götzinger, Esser does not use the horizontal brace for post-clausal subordination. Instead, he uses forward slashes like Nägelsbach (see Section 5.1.1) or Menge (see Section 5.1.2). For multiple embedding Esser uses apostrophes, similary to Rinne (see Section 3.3.4). As an example, the verse from the Nibelungenlied in (3.36 a), with an approximate translation into English in (3.36 b), is analysed by Esser as (3.36 c). The original display is shown in Figure 3.51.
(3.36) | a. | Dô die herren sâhen, daz der helt was tôt, sie leiten in ûf einen schilt, der was von golde rot, und wurden des ze râte, wie das solde ergân, daz man ez verhæle, daz es het Hagene getân. |
b. | When the men saw, that the hero was dead, they laid him on a shield, which was golden red, and it was discussed, how it should happen, that it stays hidden, that Hagen had done it. |
|
c. | (a/a’): A/b, B/a/a’/a’’ |
[3.148] Daniel Sanders (1819-1897) was a school director in Strelitz and private scholar after his shool had to close. He is most well-known for his lexicographical work (biographical details in Haß-Zumkehr 1995). He also wrote a German grammar Deutsche Sprachlehre für Volks- und Bürgerschulen (Sanders 1876), but it contains almost no analysis of sentence structure. Relevant for the topic of graphical analysis are his Deutsche Sprachbriefe (Sanders 1879), a collection of instructional texts about grammar and style addressed to a general audience. It is a collection of short explanations of grammatical phenomena, illustrated with literary examples. There are also many exercises with solutions appearing in later issues.
[3.149] Starting in the 10th issue, he introduces the concept of Satzbild (Sanders 1879: 167). Subordinate clauses are indicated by a lower-case letter corresponding to the capital for the main clause. Coordinated subordinate clauses are indicated by a number before the letter (e.g. 2a), while superscript numbers and different typefaces are used for the depth of the embedding (e.g. 𝔞2 using a small fraktur letter a). These conventions are closest to the original abbreviations introduced by Herling all the way back in 1823 (see Figure 3.6).
[3.150] Throughout the issues of the Sprachbriefe Sanders returns to these Satzbilder, adding brackets and horizontal braces. The brackets are particularly interesting, as he uses them hierarchically to enclose multiple embeddings. Horizontal braces are used to link parts of a clause that are separated by internally embedded clauses. An example of his analysis for a sentence from Goethe in (3.37 a) is repeated in (3.37 b) as shown in the original in Figure 3.52.
(3.37) | a. | Aber auch da noch wehrte sie sich so gewaltsam, dass er, um seine Augen zu erhalten und die Feindin doch nicht zu beschädigen, sein seidnes Halstuch abreißen und ihr die Hände damit auf den Rücken binden musste. (Sanders 1879: 250, citing Goethe Die wunderlichen Nachbarskinder) |
b. | Aber A so, dass (a (, um zu 𝔞2 und doch nicht 2𝔞2,) a und 2a) |
[3.151] Friedrich Blatz (1824-1900) was a teacher in Offenburg and later Oberschulinspektor in Ober-Elsaß (biographical details in Oster 1906). Towards the end of his career he wrote the massive (almost 900 pages) Neuhochdeutsche Grammatik mit Berücksichtigung der historischen Entwicklung der deutschen Sprache (Blatz 1879), only to be outdone by his third edition, which is expanded up to almost 2200 pages in two volumes, published after his retirement. The second volume contains a much expanded analysis of sentences structure (Blatz 1896). There exists also a shorter version of the book, the Neuhochdeutsche Schulgrammatik, which saw eleven editions between 1881 and 1918, but this shortened version does not contain any graphical analysis.
[3.152] In the first edition, completely at the end of the book, after the discussion of the Periode, Blatz adds a short discussion about Satzbilder (Blatz 1879: 866-871). The abbreviations used look very much like the system from Gelbe (see Section 3.3.18), especially the superscript v for Verkürzung, i.e. non-finite subordinate clauses. Blatz also offers the option to use different letters for different kinds of subordinate clauses, e.g. m for modal clauses or r for relative clauses. His use of these abbreviations is quite similar, but not identical, to the approach used by Diringer (see Section 3.3.19). Blatz does not mention either Gelbe or Diringer. An example of Blatz’ approach is shown in Figure 3.53 for the example sentence in (3.38).
(3.38) | Durch solche und ähnliche Thorheiten sind reiche und vornehme Leute an den Bettelstab gekommen und genöthigt worden, die um Hilfe anzusprehen, auf welche sie früher hochmüthig herabgesehen haben, die aber durch Fleiß und Sparsamkeit zu Vermögen sun ansehen gekommen sind. (Blatz 1879: 868, citing Benjamin Franklin Alte Goldbriefe) |
[3.153] In the major revision of the Neuhochdeutsch Grammatik published in 1896 the analysis of sentence structure is greatly expanded. The discussion of the Periode is only slightly increased, with most examples being still almost the same (Blatz 1896: 1284-1288). However, Blatz now adds a new section about the internal structure of a clause (Blatz 1896: 43-46), using the term Satzbild for both analyses. However, the internal structure of a clause looks completely different (see Figure 3.54). This analysis is probably influenced by Kern (see Section 3.3.24), whose proposals were published only after the second edition of Blatz’ book, but before the major revision. Although Blatz’ graphics look rather different from Kern’s, it is particularly telling that the verb is now alone at the top of the analysis, with all constituents depending on it, including the subject. In his explanation of this method, Blatz quite confusingly uses many different terms in a single introductory sentence:
Die Zerlegung des Satzes in seine Glieder (Auflösung, Analyse des Satzes) gründet sich auf das Durchfragen (Kontruieren) desselben. (Blatz 1896: 43)
(‘The dismantling of the sentence into its pieces (disbandment, analysis of the sentence) is based on recurrent questioning (construction) of the same.’)
[3.154] The graphical structure in Figure 3.54 is an analysis of the sentence in (3.39). The verb heißen is positioned at the top and all major constituents are listed below it. It is a bit confusing that there are no lines connecting them to the main predicate above. Further modifiers are linked with lines below the major constituents. Complete phrases are repeated, mixing constituency with dependency (to use moder terminology). This mix makes the graphical analysis really confusing to interpret.
(3.39) | Heißen Sie den Hausdiener meinen Reisesack in mein Zimmer im zweiten Stockwerk bringen. |
[3.155] Karl Zettel (1831-1904) was a teacher in Regensburg and later an editor of de-luxe editions of German poetry in Munich. He wrote a small article about the method of teaching German in schools, called Auf welche Weise kann der Unterricht in der deutschen Sprache und Literatur an unseren Studienanstalten methodisch und systematisch betrieben werden? (Zettel 1882). He argues that Satzbilder are quite useful in school, as they can readily be displayed on the blackboard (Zettel 1882: 85-87). His graphics look like those of Götzinger (see Section 3.3.1), although Zettel uses superscript numbers to indicate the parts of a clause that are separated by other clauses. Also telling is his example sentence (3.40), which is a reformulation of the example given by Bauer (see Figure 3.34).
(3.40) | Ich schenkte ihm 300 Rubel, und als ich erfuhr, dass derselbe in Kiew einen Anverwandten habe, der ihn zu sehen wünschte, stellte ich ihm frei, ob er seine Reise Fortzusetzen gedenke, oder ob er umkehren wolle. |
[3.156] Franz Kern (1830-1894) was a teacher and school director in Stettin (Szczecin) and Berlin. Most of his writings are literary commentaries for use in schools, but after his hiring in Berlin in 1881 he published various books on graphical methods for grammatical education. In the following decades, these proposals were discussed extensively in committees dealing with reforming the school curriculum, but in the end the decision went against Kern’s graphical method. More details on Kern’s life, his linguistic work and his impact can be found in Keinasto (2001; see also Osborne 2020).
[3.157] Starting in 1883, Kern published four books in quick succession. The first two were written for more learned readers, including teachers, the second two were written for a broader audience, including students. In accordance with the typesetting tradition at the time, the first two were printed in an Antiqua-typeface, in contrast to the later two, which were printed in a Fraktur-typeface:
[3.158] For his graphical approach Kern uses line diagrams. This approach is very likely inspired by Hoffmann (see Section 3.3.3), whose grammar is cited positively at various times in Kern’s work (Kern 1883a: 57; Kern 1888b: 74, 156-159). As an example of Kern’s approach, his analysis of the sentence in (3.41) is shown in Figure 3.56. These kind of line diagrams are already described in words in the Satzlehre (Kern 1883b: 110), but the first actual examples are given in the Methodik (Kern 1883a: 1-34). In his graphical analysis, Kern only deals with the dependency structure and he almost completely ignores the actual ordering of the words in the sentence. The left-right ordering of the dependents is always subject-object-adverbial, irrespective of their ordering in the sentence. Such a fixed functional ordering is reminiscent of the approach by Moritz (see Section 3.2.1), although there is no indication of any direct influence from Moritz on Kern. A completely new aspect of Kern’s diagrams is the sole placement of the finite verb on top. Earlier diagrams by other scholars always placed both the subject and the verb on top.
(3.41) | Als sie nun in ihrem neuen Glanze da stand, kamen die Sperlinge wieder, ihre alten Wohnungen zu suchen. (Kern 1883a: 13) |
[3.159] Kern calls his graphical display a Schema and at various times he compares it to a tree (Kern 1883a: iv-v, footnote; 1883b: 93, 95-96). He even explicitly discusses the option to draw the Schema from bottom to top, like a real tree, but decides against this display for practical reasons. Also note that in the following quotation Kern talks about dem gewöhnlichen Schema ‘the usual schema’, probably referring to other authors, like Hoffmann, who also use a display with modifying words positioned downwards.
Richtiger und anschaulicher (dem Bilde des Baumes entsprechend) wäre vielleicht das umgekehrte Schema, so daß das finite Verbum unten stände, und die verschiedenen Bestimmungen sich als Verästelungen und Verzweigungen darstellten; doch habe ich es nicht gewählt, weil erstens in der praktischen Durchführung die Vorausberechnung des nötigen Raumes dem Schüler Schwierigkeiten machen würde, und zweitens der herkömmliche Terminus Abhängigkeit zu dem gewöhnlichen Schema besser paßt. (Kern 1883a: iv-v, footnote)
(‘The reverse schema would maybe have been more correct and more vivid (corresponding to the image of a tree), in such a way that the finite verb would be placed at the bottom and the various modifiers would be portrayed as branches; however, I did not choose this option because, first, in the practical application the student would have had difficulties with the planning of the required space, and second, the conventional term “dependency” (Abhängigkeit) fits better to the usual schema.’)
[3.160] As for the placement of the subject, in the Satzlehre (Kern 1883b: 30-63) Kern spends many pages on the definition of the notion “subject”. He basically concludes that the subject is somehow special, but that it still depends on the finite verb just like other constituents: [Das Subjektswort im Nominativ] regiert nicht das finite Verbum, aber er ist seine allernächste Bestimmung ‘[The subject in the nominative] does not govern the finite verb, but it is its closest modifier’ (Kern 1883b: 46). Interestingly, when he explains his idea about drawing pictures, he uses a Konjunktiv when discussing the placement of the subject in the graphic (see quote below). This indicates that in late 1882, when he wrote the Satzlehre, he was not yet completely sure about the placement of the subject: either besides the finite verb (just like Hoffmann, see Section 3.3.3), or below the finite verb (which he chose eventually). Notably, this Konjunktiv is changed to a regular indicative in the second edition of the Satzlehre (Kern 1888b: 177), as by then he is committed to the finite verb as the sole head of the clause with the subject placed below it. A verb-centric display was already used ten years earlier by Brassai (cf. Section 5.5.2), but there is no direct evidence of any influence from Brassai on Kern. Still, Kern’s graphics are eerily similar to Brassai’s.
Dem Sachverhalt noch entsprechender freilich wäre es, wenn oben ganz allein stets Verbum finitum stände und auch das Subjektswort am Ende eines nach links unten geführten Striches stände als die Verbalperson bestimmend. (Kern 1883b: 110)
(‘It would be even more in accordance to the issue, if the finite verb would be placed alone on top and the subject would also be placed at the end of a line going down left as the determiner of the verbal person.’)
[3.161] With simple examples, Kern’s approach is a real step forward. However, the more examples he discusses, the more confusing his whole systems becomes. The first problematic aspect is that in the Grundriss (Kern 1884) he starts leaving out the actual words in his schemas. He mostly just displays the grammatical categories, using constantly changing abbreviations. Such an “empty” schema is called a Satzbild (Kern 1884: 30, footnote). It is left up to the reader to reconstruct which abbreviations in the graphic display correspond to which words in the example sentence. This task is often non-trivial, because the ordering in the Schema does not necessarily represent the order in the sentence.
[3.162] Second, Kern starts to use dotted lines without any explanation. At first this seems to be restriced to coordination (Kern 1884: 31), but then he also uses dotted lines to indicate levels of embedding in complex sentences (Kern 1884: 56). In the Leitfaden (Kern 1888a) he starts using dotted lines everywhere between the levels of modification. Instead of clarifying the structure, these dotted lines are mostly just adding confusion.
[3.163] Third, for the analysis of complex sentences with many subordinated clauses, Kern introduces additional square brackets and round brackets, as shown in Figure 3.57 (Kern 1884: 56-57). This graphic is extremely confusing and it indicates that Kern’s approach reaches its limits with more complex sentences. This graphic is an analysis of the complex sentences in (3.42).
(3.42) | Wenn dein Bruder, der mir so wert und teuer ist, wie nur irgend ein Mensch es sein kann, in der Not, die ihn bedrängte, sich an mich gewendet hätte, so würde ich, was nur irgend in meinen Kräften stand, aufgeboten haben, damit er damals die Ruhe, deren er so würdig ist, wieder gewonnen hätte. (Kern 1884: 56) |
[3.164] Instead of visualising all modifications in all clauses, The display in Figure 3.57 focusses on the relationship between the various subordinate clauses, similar to the tradition of the Periodenbilder. Unfortunately, Kern also includes some non-clausal constituents, but not all. For example, the head nouns of relative clauses are included. These few non-clausal constituents are placed inside round brackets and these elements in round brackets then are linked to other parts of their subordinate sentence by using square brackets.
[3.165] The result is confusing enough that even Kern himself missed a printing error in this graphic (from the text it is clear he correctly understood the sentence structure): on subordination level 2 (N-S. 2) the element Präd.꞊N (=predicative nominative) belongs to the Attr.꞊S. (=attributive clause) to its left (der mir so wert und teuer ist), not to the one on its right (die ihn bedrängte). So, the square brackets should have been placed around the first and the second element, not around the second and the third.
[3.166] The graphical analysis would have been much more transparent when only the subordinate clauses would have been shown. The internal structure of each of these clauses could, for example, have been explained separately as individual schemes. Figure 3.58 shows an attempt to “clean up” Kern’s analysis from Figure 3.57.
[3.167] Kern himself also realised that his system was not ideal for large sentences with many subordinate clauses. Consequently, he added the option for a simplified Satzbild only consisting of letters and numbers, like the many proposals of Periodenbilder ever since Herling. His simplified analysis for (3.42) is shown in Figure 3.59. It is a very simple one-line Periodenbild with a completely idiosyncratic use of letters and numbers. The dots separate the clauses. Each subordinate clause is represented by a number, which indicates the depth of the embedding. Lower-case letters a, b are used to indicate that a clause is separated by other clauses, so e.g. all parts labelled 1a are parts of the same subordinate clause. Oddly, the two parts of the main clause, labeled H (=Hauptsatz), are not marked in this way, probably because Kern assumes there to be only a single main clause in such a simplified analysis.
[3.168] Finally, the following two details from Kern’s descriptions suggest influences that might have led Kern to his particular graphical approach. First, he argues against the term Erweiterung with an argumentation that is eerily similar to the one made by Diesterweg as discussed in Section 3.2.6 (Diesterweg 1830: 167; 1834: 151). So maybe Kern knew about Diesterweg’s approach, although there is no other concrete evidence for this.
Uebrigens besagt der Ausdruck Satzerweiterung gar nichts als Satzverlängerung; denn durch hinzugefügte Bestimmungen wird der Inhalt des Satzes natürlich nicht erweitert, sondern verengert und dadurch bereichert. (Kern 1883a: 5, footnote)
(‘Incidentally, the term Satzerweiterung ’sentence expansion’ refers to nothing more than elongation of the sentence; however, by adding modifications the content of a sentence is not expanded, but narrowed and enriched in the process.’)
[3.169] Second, Kern does not like the notion of bekleidetet ‘dressed’ sentences. His argument is actually quite convincing: this metaphor suggest that parts of the sentence are like shirts or trousers, which is actually quite a strange metaphor. Yet, even more telling is Kern’s mentioning of the alternative term umkleidet, which is a very specific term only used by Hoffmann (1839: 146-148).
Eine recht überflüssige doch weniger schädliche Terminologie, ist auch die des nackten und bekleideten (oder umkleideten) Satzes. … In dieser Terminologie wird es vorgezogen, die Bestimmungen, statt als Äste, Zweige und Blätter sich zu denken, sie sich lieber als Hemde, Rock und Weste vorzustellen. Ich finde dieses Bild weder treffend noch geschmackvoll. (Kern 1883b: 95-96)
(‘A rather superfluous, though less harmfull terminology, is the one of nackte ’naked’ and bekleidete ‘dressed’ (or umkleidete ‘around-dressed’) sentences. … In this terminology modifiers, instead of thinking about them as branches and leaves, are preferably imagined as shirt, skirt and waistcoat. I consider this image to be neither fitting nor tasteful.’)
[3.170] Franz Willomitzer (1847-1910) was a teacher for German language and history in Vienna, Austria. His book Deutsche Grammatik für österreichische Mittelschulen first appeared in 1879 and was revised and reprinted many times up to a 23th edition in 1930. After the death of Willomitzer the book was further developed by Johann Tschinkel. He renamed the book to Deutsche Sprachlehre für Mittelschulen and strongly reduced the number of visualisations, with eventually only a single unexplained graphic remaining (Tschinkel 1921: 242).
[3.171] In the first edition Willomitzer uses a numbered table to analyse a sentence, in the style of Becker (see Section 3.2.5). The analysis of the sentence Finsternis bedeckt die Erde ‘darkness envelops the earth’ is shown in Figure 3.60 (Willomitzer 1879: 123). Just like Becker, Willomitzer uses reference to the row-numbers to clarify the dependency structure. However, different from Becker, Willomitzer does not add any dependency between subject and predicate (similar to Fürg, see Section 3.3.7). Unfortunately, this example sentence is actually way too simple to really clarify the intentions of Willomitzer, and there are no other, more complex examples presented of this analytic approach. Exactly the same figure is repeated in later editions (Willomitzer 1882: 109; 1885: 122; 1894: 113).
[3.172] For sentences with multiple subordination like in (3.43) Willomitzer uses a different visualisation as shown in Figure 3.61 (Willomitzer 1879: 157-158). Each clause is indicated by a horizontal line with a letter, and these clauses are connected by slanted lines. A line left-downwards indicates a preposed subordinate clauses, while slanted lines right-downwards indicate postposed subordination. Internal subordination, like with relative clauses, is not discussed. Exactly the same visualisation is also used in the fourth edition (Willomitzer 1882: 144-145).
(3.43) | a. | Als Karl der Große den Kaiserstuhl Roms bestieg, war ein Ziel erreicht, dem hochstrebenden deutschen Fürsten seit Jahrhunderten nachgetrachtet haben. |
b. | Nur der ist unser Wahrer Freund, der mit uns gleiche Bestrebungen teilt und auf den wir uns in guten und bösen Tagen verlassen können. | |
c. | Ich erinnere mich, dass wir in einem alten Haus wohnten, welches eigentlich aus zwei durchbrochenen Häusern bestand. |
[3.173] There are a few interesting changes in the sixth edition (Willomitzer 1894: 154-156). First, Willomitzer adds a “plus” symbol for coordination instead of the dotted connection in the middle image of Figure 3.61. Second, he adds one-line Periodenbilder as an additional option, although he does not explain this method at all. He simply adds abbreviations to the end of a few examples sentences, as shown in Figure 3.62. The format of these abbreviations is similar to other approaches before him, except for the use of the colon and the forward slash. Willomitzer reverses the use of the colon and the forward slash in comparison to e.g. Nägelsbach (cf. Figure 5.1). Willomitzer uses the forward slash to separate preposed subordination, while the colon separates postposed subordination.
[3.174] Willomitzer adds a new kind of graphical display in the revision between the third edition (Willomitzer 1882) and the fourth edition (Willomitzer 1885). In this timeframe Kern published his new graphical approach (see Section 3.3.24) and Willomitzer immediately includes Kern’s approach in his fourth edition, though without explicit attribution (Willomitzer 1885: 123, 132, 137). The examples below are cited from the almost identical sixth edition because of the better quality of the available scans (Willomitzer 1894: 114, 121, 127). The analysis of the first example (3.44 a) is shown in Figure 3.63. This graphical display is much clearer than the examples from Kern, mostly because Willomitzer does not use any abbreviations. The analysis of (3.44 b) is shown in Figure 3.64. This is an example with a subordinate clause, in which the complementizer dass is not a node in the graph, but a relation between nodes, an approach also used by Kern.
(3.44) | a. | Der Biss der in Amerika einheimischen Klapperschlange tödtet in wenigen Augenblicken den kräftigsten Menschen. |
b. | Wir fordern, dass jeder seine Pflicht thue. |
[3.175] Otto Lyon (1853-1912) was a teacher and Stadtschulrat in Dresden (biographical details in Kolb 1987). He was an extremely productive author, revising and reissuing classical linguistic works of the likes of Heyse, Becker and Eberhard, editing an anthology of German literature for use in schools, and producing a textbook for teaching all aspects of German. He also co-founded the journal Zeitschrift für den deutschen Unterricht.
[3.176] His textbook Handbuch der deutschen Sprache für höhere Schulen was originally published in 1885. There have been many different versions of this handbook. I have counted at least 36 editions among seven different variants, some having multiple volumes, including six different co-authors. These books are often indexed incompletely or wrongly in catalogues, so it takes some effort to identify which version is available at any library. I have tried to reconstruct the editorial developments, but this edition-mess is in need of a more proper clean-up. Different variants of the handbook were introduced starting with the sixth edition in 1902. The second volume of Lyon’s original sixth edition from 1902 is available online at https://google.com/books?id=ZQkYT08lhVYC. A distinction into four different versions was made, which are sometimes (but not always) indicated by the letters A through D. I will use these letters as a shortcut to identify the different versions, and I will add a subscript for major revisions of the C and D variants.
[3.177] Already in the first edition, Lyon proposes a set of abbreviations for different syntactic functions within the clause (Lyon 1889: 99-100). Unfortunately, these abbreviations are completely opaque and they have almost no mnemonic value. For example, subject is called “I”, the predicate is “II”, the direct object is “IIIa”, a causal adverbial is “IVd”, etc. These abbreviations are used to analyse a sentence. The example sentence (3.45) is used in almost all variants to illustrate the process. The names for these methods and their details differ between the variants, as shown in Figure 3.65 for the A and the D1 variant (called Satzanalyse), Figure 3.66 for the D2-variant (called schriftliche Satzzerlegung) and Figure 3.67 for the C1-variant (called Satzbild).
(3.45) | Der Löwe schenkte im Walde der bittenden Maus großmütig das Leben. |
[3.178] New graphical approaches were introduced in the 1910s, both in the C2 and in the D2 variants. Both variants introduce line diagrams, but with some interesting differences. The D2-variant introduces a verb-centric diagram as shown for example (3.46 a) in Figure 3.68. This graphic is very similar to Kern’s approach (see Section 3.3.24). Kern is not explicitly mentioned, but in the preface the authors mention that: wir [sind] wie die meisten neueren Sprachmethodiker vom Prädikat ausgegangen ‘just like most newer linguistic methods, we started with the predicate’ (Müller & Roedel 1920: iii).
[3.179] The line diagram in the C2-variant does not have a verb-centric approach, but instead places subject and finite verb equally on top, cf. Figure 3.69 for the example sentence (3.46 b). Both the C2 and the D2-variant add syntactic labels to all words in the line diagrams. In both cases, these labels are abbreviations that consists of two parts: an indication of word class with a lower-case letter and a symbol for the syntactic function using capitalised abbreviations. However, the naming conventions are completely different as explained in the figure-captions.
(3.46) | a. | Die strahlende Sonne leuchtet am wolkenlosen Himmel. |
b. | Der fleißige Schmied hämmert das glühende Eisen auf dem harten Amboß. |
[3.180] In the C2-variant there is a fascinating application of the line diagram to explain a syntactic detail of German (Sieke & Reißig 1919: 136). The authors observe that with the lassen+Infinitiv and the heißen+Infinitiv construction in German there is a crucial difference between using an accusative or a dative. Their examples are repeated here in (3.47) with the crucial parts italicised. When accusative, the participant is the patient of the finite verb lassen/heißen, while when dative, this participant is the recipient of the infinitive verb vorlesen/bringen. This difference is illustrated by the contrasting graphical syntactic analyses in Figure 3.70.
(3.47) | a. | Mein Freund ließ michakk die Zeitung vorlesen. Mein Freund ließ mirdat die Zeitung vorlesen. |
b. | Der Graf hieß den Knechtakk das Roß bringen. Der Graf hieß dem Knechtdat das Roß bringen. |
[3.181] Separate from the clause-internal analysis, Lyon also discusses Satzbilder in the tradition of Lehmann (see Section 3.2.8) for sentences consisting of multiple clauses. Lyon presents a very short one-page summary of his notational variant (Lyon 1889: 269-270). He uses different lower-case letters for different kinds of subordinate clauses (a=complement clause, b=relative clause, c=adverbial clause), round brackets for non-finite clauses and superscript numbers for levels of embedding. The punctuation of the original sentence is repeated verbatim in the Satzbild. As an example he presents the graphical analysis in Figure 3.71 for the sentence from Schiller in (3.48).
[3.182] Exactly the same explanations and examples are found in the D1-variant (Lyon & Scheel 1911: 164-165), the D2-variant (Müller & Roedel 1920: 268) and the C2-variant (Sieke & Reißig 1919: 197-198). Only in the C1-variant there is a short note offering an alternative notation (Lyon & Polack 1902: 154-156). In this variation, a superscript v is used for non-finite clauses (cf. Gelbe, Section 3.3.18) and a fractional number is used for parts of clauses that are separated by another clause (cf. Herling, Section 3.2.2).
(3.48) | Obgleich er in seinen Anstalten weit genug vorwärts gerückt war, um die Stadt zu beängstigen, so wared doch noch sehr viele Schritte zu thun, um sich wirklich zum Meister derselben zu machen. (Lyon 1889: 270, citing an example from Schiller Geschichte des Abfalls der vereinigten Niederlande) |
[3.183] Lyon also wrote a separate series of anthologies of German literature called Die Lektüre als Grundlage eines einheitlichen und naturgemäßen Unterrichtes in der deutschen Sprache. In volume 2, part 1 (for Obertertia, i.e roughly for teenagers of age 15) there is an extensive discussion of the use of these Satzbilder with numerous examples (Lyon 1897: 130-154). To finish the discussion of Otto Lyon’s graphical analysis, I have repeated here in (3.49) one of Lyon’s examples from Lessing with the corresponding Satzbild in Figure 3.72. The reader is invited to follow along, just like the teenage students were supposed to do back at the later part of the 19th Century.
(3.49) | Es ist wahr, mit dergleichen leidigen Nachahmungen fängt das Genie an zu lernen; es sind seine Vorübungen; auch braucht es sie in größeren Werken zu Füllungen, zu Ruhepunkten unserer wärmeren Teilnehmung: allein mit der Anlage und Ausbildung seiner Hauptcharaktere verbindet es weitere und größere Absichten: die Absicht, uns zu unterrichten, was wir zu tun oder zu lassen haben, die Absicht, uns mit den eigentlichen Merkmalen des Guten und Bösen, des Anständigen und Lächerlichen bekannt zu machen; die Absicht, uns jenes in allen seinen Verbindungen und Folgen als schön und als glücklich selbst im Unglücke, dieses hingegen als häßlich und unglücklich selbst im Glücke zu zeigen, die Absicht, bei Vorwürfen, wo keine unmittelbare Nacheiferung, keine unmittelbare Abschreckung für uns statthat, wenigstens unsere Begehrungs- und Verabscheuungskräfte mit solchen Gegenständen zu beschäftigen, die es zu sein verdienen, und diese Gegenstände jederzeit in ihr wahres Licht zu stellen, damit uns kein falscher Tag verführt, was wir begehren sollten, zu verabscheuen, und was wir verabscheuen sollten, zu begehren. (Lyon 1897: 140, citing Lessing’s Hamburgische Dramaturgie) |
[3.184] Heinrich Utzinger (1842-1913) was a teacher and later schooldirector in Küsnacht, Switzerland. He wrote the book Deutsche Grammatik: Lehr- und Übungsbuch für Sekundarschulen (Utzinger 1887), which was written on behalf of the kanton Zürich to be the official textbook for all secondary schools in the kanton. Utzinger himself delivered various editions during his lifetime, though with only minor changes. The book remained in print after his death up to a 14th edition in 1943. In the preface he mentions various grammars as influences, but none of those could have been the inspiration for his graphical display.
[3.185] In the discussion of word order in the clause, Utzinger presents a single illustrative Schema for the example sentence in (3.50) as shown in Figure 3.73 (Utzinger 1887: 89). The word Schema and the use of slanted lines might indicate an influence of Kern (see Section 3.3.24). However, Utzinger follows the old tradition in placing subject and predicate side-by-side on top of the graphical display. Because there is only a single example of this graphical approach, it remains unclear how Utzinger envisions the left-right ordering of modifiers to be used. Kern quite radically used a strict subject-object-adverbial order, but the single example from Utzinger does not clarify whether he would follow that approach.
(3.50) | Die älteste Geschichte eines jeden Volkes ist gewöhnlich durch mancherlei Sagen verdunkelt. |
[3.186] For the analysis of complex sentences with multiple subordinate clauses Utzinger proposes a different format (Utzinger 1887: 122). He uses a one-line summary, which he calls a Satzbild, as shown in Figure 3.74 for the examples sentence in (3.51). The depth of embedding is indicated by subscript numbers. Lower-case letters indicate subordinate clauses, but it remains unclear why the letters a and b are used multiple times in this example. In the example, the two parts marked a1 are actually two separated parts of the same subordinate clause, but all other parts are separate clauses, and might better have been indicated by separate letters.
(3.51) | Als der Graf von Strassberg, welcher über den Brünig ins Unterwaldnerland eingefallen war, von der Niederlage des Herzogs Leopold Kunde erhilet, kehrte er schleunig zurück, da er wohl einsah, dass ein weiteres Vordringen unnütz sein würde. |
[3.187] Béla Kuderna (1846-1915) was a captian (Hauptmann) in the Austrian army and teacher for German language and style at the Infanterie-Cadetenschule in Vienna, Austria. In 1893 he wrote a short article on teaching German, called Sprachliches Mahnwort (Kuderna 1893). This article includes two mostly unexplained examples of Satzbilder as he used them in his classes (Kuderna 1893: 268-269). Shortly afterwards he published a complete booklet about his approach to graphically analyse sentences, called Das Satzbild in seiner Anwendung für die Lehre vom Satze (Kuderna 1895). This book goes into great detail explaining his approach, including numerous practical examples.
[3.188] As for the inspiration for his graphical approach, Kuderna mentions Bauer (see Section 3.3.8) in passing (Kuderna 1895: 4), but only to complain about Bauer’s non-mnemonic usage of letters and symbols. For example, instead of using an abstract letter A for the main clause, Kuderna proposes to use the letter H as an abbreviation of the German word Hauptsatz ‘main clause’. There are no further clear indications for specific influences. Kuderna uses line diagrams and letters with superscripts and subscripts, just like most other authors around this time. However, the combination of line-diagram with abbreviations as nodes is an idiosyncratic mixture of Kuderna. Subject and predicate are both placed on top, indicating a more traditional approach, not following Kern’s verb-centric proposal (cf. Section 3.3.24).
[3.189] Kuderna’s Satzbild comes in two different variants, one for single clauses (using lower-case letters) and one for complex sentences consisting of multiple clauses (using capital letters). To explain his system, Kuderna presents the two example sentences in (3.52), graphically analysed in Figure 3.75. Sentence (3.52 a) is an example of a single clause and (3.52 b) is a complex sentence with multiple clauses. Kuderna’s diagrams are shown to the left in Figure 3.75 (using letters). The diagrams to the right with the actual words are only added as explanation. This is the only time in the book that show the actual words, all other examples only have the abstract letters.
(3.52) | a. | Die grausame Wirklichkeit zerstört oft unsere schönsten Hoffnungen. |
b. | Es ist natürlich, dass man wahrhaft große Männer bewundert und verehrt; den sie heiligen das Volk zu dem sie gehören. |
[3.190] Kuderna extols his own system because of the transparent usage of the letters (s=Subjekt, a=Attribut, o=Objekt, H=Hauptsatz, etc.). However, he complicates his analysis by adding superscripts and subscript, ending up with more than 50 different abbreviations (summarised in Kuderna 1895: 10-11). As a result, more involved examples very quickly become unwieldy and difficult to interpret. For example, the sentence in (3.53) are analysed by Kuderna as shown in Figure 3.76 and Figure 3.77. Instead of simplifying the interpretation of the sentence structure, these Satzbilder become puzzles in their own right.
(3.53) | a. | Finsteren Blickes, aber erhobenen Hauptes folgte die lange Reihe der Gefangenen dem mit allen erdenklichen Zeichen von Pracht und Glanz geschmückten Wagen des Imperators bis zu dem weiten, jetzt dicht mit lärmendem Volke besetzten Festplatz des Forums. (Kuderna 1895: 33) |
b. | Die Bundesgenossen, welche dazu bestimmt waren, an dem Kampfe unmittelbar theilzunehmen, und denen man deshalb besondere Führer zugeweisen hatte, erhielten die Weisung, sich in der Hafenstadt zu sammeln, um von da, wenn alles beisammen wäre, nach den bedrohten Südprovinzen eingeschifft zu werden. (Kuderna 1895: 52) |
[3.191] Hans Trunk (1847-1929) was a teacher and school director in Graz (Austria). I have not been able to find much biographical information about Hans Trunk. HathiTrust provides a birthyear (https://catalog.hathitrust.org/Record/102143919) and Kürschners Deutscher Literatur-Kalender (1930: 1524) lists his death, but without precise date (https://doi.org/10.1515/9783112340301). Most of his writings deal with pedagogical methods. In this context he writes a small article Über Satzbilder (Trunk 1895). Much of the content of this article is also included in a more wide-ranging book about teaching German language Zur Hebung des deutschen Sprachunterrichtes (Trunk 1898: 123-127).
[3.192] Trunk’s graphical analysis is inspired by Kern (see Section 3.3.24), who is explicitly cited (Trunk 1895: 309; Trunk 1898: 123). Additionally, Trunk clearly knew about many other approaches as well, citing the historical survey of Satzbilder by Gelbe (1880). However, Trunk complains about missing simplicity and clarity of all previous proposals for Satzbilder, so he devised his own variant. Although he mostly based his approach on Kern (using line-diagrams with the verb as the sole governor of the clause), he adds the requirement that all leaves of the graph should be placed in the order as they occur in the sentence (Trunk 1895: 310). Specifically, the direction of the slanted lines in this graphical display is meaningful: slanted down left means the element comes before its head, while slanted down right indicates placement after the head.
[3.193] Just like Kern, Trunk uses code-letters in his graphical displays, which makes it actually quite hard to interpret them. This is particularly troubling when many elements with the same code co-occur in the analysis, as in Figure 3.78 for sentence (3.54 a). Observing this problem, Trunk suggests to write the words themselves on the board for clarity (Trunk 1895: 314). Also just like Kern, Trunk uses a different approach for sentences consisting of multiple clauses, as in Figure 3.79 for sentence (3.54 b). Here each letter indicates a clause, not an individual word.
(3.54) | a. | Gestern schlug der Blitz in die alte Linde vor dem Hause unseres Nachbarns. (Trunk 1895: 312) |
b. | Wenn dein Bruder, der mir so wert und teuer ist, wie nur irgend ein Mensch es sein kann, in der Not, die ihn bedrängte, sich an mich gewendet hätte, so würde ich, was nur irgend in meinen Kräften stand, aufgeboten haben, damit er damals die Ruhe, deren er so würdig ist, wieder gewonnen hätte. (Trunk 1895: 314; citing Kern 1884: 56) |
[3.194] Matija (Matthias) Heric (1855-1927) was a teacher in Neukirchen (Novi Cerkvi) near Pettau (Ptuj) in the region Steiermark (today part of Slovenia). More biographical details are available in Šlebinger (1926). He wrote a short article expanding upon Trunk’s approach (see Section 3.3.29), specifically discussion nominal predication. Heric used the same graphical approach also in a short article about his native Slovene, using the name obrazec ‘formula’ (plural obrazci) for the graphical display (Heric 1897). As an example of Heric’s approach, the analysis for his examples in (3.55) is shown in Figure 3.80.
(3.55) | a. | Diese Blume wird Schneeglöckchen genannt. |
b. | Wir nennen diese Blume Schneeglöckchen. |
[3.195] Otto Toifel (dates unknown) was a teacher in Austria, first in Salzburg and later in Ried im Innkreis. He wrote a few articles in which he applied Lehmann’s graphical approach (see Section 3.2.8) to analyse examples from Middle High German literature. In particular, Toifel used one-line Satzbilder following Lehmann (1856), whom he explicitly cites. However, Toifel employs lower-case latin letters with superscript numbers for levels of subordination, instead of the different scripts used by Lehmann.
[3.196] Toifel applies this approach in two short articles to analyse “unusual” sentence structures in the German epic Kudrun (Toifel 1896; Toifel 1897). A few years later he did the same for special constructions in texts by Bertold of Regensburg (Toifel 1901). An example of his analysis is the Satzbild shown in Figure 3.81 for the quotation from the Kudrun in (3.56). The symbols in the example are not in Toifel’s original, but added here for comparison with the Satzbild. An attempted English translation is given in (3.57). The special phenomenon highlighted by Toifel is the relative clause b3, which he analyses as a preposed relative clause, refering to die ‘them’ in the following clause b2.
[3.197] In his graphical analysis, Toifel uses the colon in a special way. Lehmann originally used the colon only to indicate subordinate clauses that are preposed to the main clause (i.e. clauses in the German Vorfeld like a1 ). Toifel additionally uses the colon for subordinate clauses that are preposed to another subordinate clause, which is a highly unusual structure in German. Specifically, the fact that clause b3 is subordinated and preposed to clause b2 is indicated by the colon between these clauses in Figure 3.81.
(3.56) | (a1) und haete ich hundert swester, (A) die lieze ich sterben ê, (b1) dann ich mich alsô starke in vremeden landen haele, (b3) die man mir nam mit sturme, (b2) daz ich die minen grimmen vinden staele. (Toifel 1896: 19, citing Kudrun verse 1256, abbreviations added) |
(3.57) | (a1) and even if I had a hundred sisters, (A) I would rather let them die, (b1) before I would so strong in strange lands hide, (b3) the ones that were taken from me in battle, (b2) to steal them from my furious enemies. |
[3.198] Adolf Stamm (born 1857) was a teacher in Iserlohn and later a school director in Anklam and in Mülheim an der Ruhr. I have not been able to find more biographical information about him, except for the footnote on page 11 of this school report https://nbn-resolving.org/urn:nbn:de:hbz:061:1-899851. He is mentioned as the first director of the Gymnasium in Mülheim on https://de.wikipedia.org/wiki/Otto-Pankok-Schule. Both websites accessed 9 January 2025. He wrote a small booklet called Graphische Darstellung der deutschen Satzlehre nebst einer Interpunktionslehre (Stamm 1899) in which he presents his own, completely idiosyncratic, system of graphical sentence analysis. The scholars who took notice of Stamm’s work appear to be quite critical (Arens 1902; Eberhard 1905; Mossner 1960), mostly because his images are too complex to be practical. Actually, Stamm’s system of visual analysis is beautiful, well though-through and explained in much detail with very many examples. But indeed, it is not practical. It is very hard to remember what all symbols mean because they are completely arbitrary representations of linguistic structure.
[3.199] In the preface, Stamm mentions that the idea to use graphical display is not new, but he finds the other proposals not satisfying (Stamm 1899: v). He explicitly mentions the Handbuch of Lyon (see Section 3.3.26), but at this point in time that Handbuch did not yet include graphical analyses. So it remains unclear what other graphical approaches have influenced Stamm.
[3.200] Stamm proposes different systems for the analysis of single-clause sentences and for multi-clause complex sentences. The images for single-clause sentences at first seem reasonably easy to grasp, as illustrated here with example (3.58 a) displayed in Figure 3.82. They follow the traditional subject-predicate basis with various elements “growing” upwards out of this basis. However, there are different kinds of squiggly lines representing different kinds of structures, with numbers and letters arbitrary assigned to various functions. The same numbers even have different meanings depending upon which squiggly line they are used with.
[3.201] For a multi-clause sentences (Satzgefüge) Stamm uses a variant of the music-score metaphor, illustrated here with example (3.58 b) displayed in Figure 3.83. The vertical line representes the main clause, and all subordinated clauses are indicated by horizontal lines, appearing in the order as they occur in the sentence. The height of the horizontal clauses represents the depth of the embedding. Different letters and dashes all have different meaning, which again are assigned completely arbitrarily. For example, the double dash for clause e3̱̱ indicates that it is a conditional clause, and the double vertical dashes with clauses d2 and e4 indicate that these are non-finite participle clauses.
(3.58) | a. | Das neue Haus meines lieben endlich zurückgekehrten Freundes liegt im Walde. (Stamm 1899: 7) |
b. | Es haut nach mir mit grimmen Zähnen, c2 als meine Hunde, d2 wutentbrannt, c2 an seinen Bauch mit grimmigen Bissen sich warfen, e3̱̱ dass es heulend stand, e4 von ungeheueren Schmerz zerissen. (Stamm 1899: 17, citing Schiller Der Kampf mit dem Drachen) |
[3.202] Karl Jauker (1850-1924?) was director of the teacher’s seminary in Graz, Austria. Karl Jauker is regularly included in the Kürschners Deutscher Literatur-Kalender. The last, very short, mention of is from 1924 (page 426) https://doi.org/10.1515/9783111438030. Afterwards he is removed, and there is no announcement of his passing. He wrote a German textbook called Deutsche Sprachlehre, which appeared in various version for different audiences. I have been able to access the version für österreichische Bürgerschulen (Jauker 1900) and the version für Lehrer- und Lehrerinnen-Bildungsanstalten. This second version was first published in 1904, but it is cited here from the third, mostly unchanged, edition (Jauker 1910). More in-depth research is needed to compare Jauker’s various textbooks and their different editions.
[3.203] The visual analysis of sentences, called Zergliederung or Analyse (Jauker 1900: 78-81, 108-111; Jauker 1910: 96-99, 129-138) is clearly influenced by Trunk (see Section 3.3.29), who also lived and worked in Graz. Jauker uses different visualisations for (i) the internal structure of single-clause sentences and (ii) the clausal hierarchical structure of multi-clause sentences. In both kinds of images he follows Trunk in including sentence ordering of a dependent in relation to its immediate head: lines going down-left means that the dependent is placed before its head, while lines going down-right indicate placement after the head.
[3.204] Jauker’s Satzbild for single-clause sentences is illustrated here in Figure 3.84 for the example in (3.59 a), from the version für österreichische Bürgerschulen (Jauker 1900: 80). The different parts of the sentence are marked with abbreviations for syntactic functions. The placement of the consituents follow the ordering of the sentence, except for the attributive modifiers, called Beif=Beifügung, which are placed exactly below their heads. A few years later, in the version für Lehrer- und Lehrerinnen-Bildungsanstalten (Jauker 1910), there are two minor changes to this display. First, as shown in Figure 3.85 for the example (3.59 b), the figure only includes the functional abbreviations, so it is not immediately obvious how this figure relates to the example sentence (this was earlier presented as an alternative option). Additionally, the attributive modifiers Beif=Beifügung are now placed in such a way as to indicate their ordering relative to their heads..
[3.205] Because of the omission of the actual words in Figure 3.85, it becomes a syntactic exercise to related the figure tot the actual example. In this figure, the clause is headed by the finite verb stand ‘stood’, which is indicated as Ausſag. Ztw.=aussagendes Zeitwort. Relative to this verb, there is one line going down-left to the preposed adverbial phrase headed by Stelle ‘place’, which is indicated as U.d.O.=Umstand des Ortes. To the right is the postposed subject Menge ‘large number’, which is indicated by Subjw=Subjektwort. Both the adverbial phrase and the subject have various attributive modifiers, called Beif=Beifügung.
(3.59) | a. | In einem strengen Winter trieb der Hunger ein ganz erschöpftes Rothkehlchen an das Fenster eines frommen Landmannes. |
b. | An einer sonnigen Stelle inmitten hoher schlanker Buchen stand eine Menge weißer Anemonen und gelber Schlüsselblumen. |
[3.206] The Satzbild for a multi-clause sentence only indicates the various clauses in the graphic, as illustrated in Figure 3.86 for the sentence (3.60 a) and in the highly similar Figure 3.87 for the sentence (3.60 b). These examples come from different versions and use different example sentences, but they have almost the same structure. Probably Jauker changed the example to appeal to the audience of these books, the first being for Bürgerschulen, the second for the students at his teacher’s seminary.
[3.207] The capital letter H indicates the main clause, while subordinated clauses are indicated by lower-case letters. The depth of embedding is indicated by superscript numbers. The rounded arches indicate that the linked abbreviations are actually two parts of the same clause, interrupted by a subordinate clause. There are various different abbreviations for the syntactic function of the subordinate clause in these figures, e.g. a manner adverbial clause uw=Umstand der Weise, a causal adverbial clause ug=Umstand des Grundes, an object clause o=Objektsatz, various relative clauses b=Beifügungssatz. The round brackets for the clause (ug1) indicate that this clause is non-finite.
(3.60) | a. | uw1 Wie mehrere Fensterscheiben, b2 die einzeln vollkommen durchsichtig erscheinen, uw1 eine grünlich-blaue Färbung haben, ub1 wenn man sie aufeinanderlegt, H so zeigt auch die atmosphärische Luft, b1 die in dünnen Schichten farblos ist, H ein schönes Blau, ub1 wenn wir durch größere Strecken hindurch sehen. |
b. | uz1 Als der römische Stadtpräfekt, b2 der später Papst Gregor I. wurde, uz1 auf dem Sklavenmarkte Knaben aus Angeln aufgestellt sah, b2 die ein Händler eingeführt hatte, H fragte er, (ug1) von den blonden Locken, den weißen Leibern und den holden Kindergesichtern ergriffen: o1 “Woher sind sie gebracht?” |
[3.208] Siegmund von Raumer (1860-1939) was a teacher in Erlangen. He wrote a short article about Periodenbilder complaining about their complexity and unsuitability for usage in education. He is particularly critical of Kern (cf. Section 3.3.24), discussing the example shown earlier in Figure 3.57. Just like Kern himself, Raumer also misses that there is an error in that graphic.
[3.209] As a summary of his criticism Raumer presents various renditions of Periodenbilder for the same complex sentence, repeated here in (3.61). Two different proposals by Kern are illustrated, alongside displays in the style of Nägelsbach (cf. Section 5.1.1) and Götzinger (cf. Section 3.3.1). I have not been able to find the source of proposal number (3) in Figure 3.88. There is no citation in the article itself, and none of the proposals currently discussed in this book look like this Periodenbild.
[3.210] The last proposal (6) is the one preferred by Raumer. He argues for a simple illustration, that is easy to grasp and might actually help to understand the sentence. Each clause is given a different letter, with main clauses in capitals and subordinate clauses in lower-case letters. Multiple occurrences of the same letter indicate that the clause is separated into parts by another clause. The left-right ordering follows the order of the sentence, while the top-down ordering indicates the depth of the embedding. In effect, Raumer’s proposal is very close the the various music-score visualisation for complex sentences, for example by Diesterweg (see Figure 3.15) or Schlecht (see Figure 3.40). However, Raumer does not explicitly cite any of these earlier proposals.
(3.61) | Bei der grossen Trennung, welche das Evangelium Jesu, sobald es gepredigt wurde, zwischen seinen Anhängern und zwischen Juden und Heiden hervorbrachte, mussten unzählige Handlungen, die man sonst für erlaubt oder wohl gar für pflichmässig hielt, zweideutig und verdächtig werden. (Raumer 1902: 63, citing an example from Götzinger) |
[3.211] Otto Ludwig Theodor Mensing (1868-1939) was a teacher in Kiel and wrote extensively about Low German grammar. He is most well-known for his Schleswig-Holsteinische Wörterbuch (more biographical information is available in Molzow 2000). He also wrote various different textbooks for teaching German language and literature, among them the Deutsche Grammatik für höhere Schulen. His textbooks were very popular and appeared in various editions up to 1965. I have tried to access many different versions, but as Molzow (2000: 1759) notes, textbooks are often treated as consumer goods and are not always properly archived. The many versions and editions of Mensing’s textbooks needs a more detailed investigation than I am able to provide here.
[3.212] Mensing originally wrote various separate textbooks, but around 1915 they are combined into a series, called Hilfsbuch für den deutschen Unterricht auf höheren Schulen. As for the Deutsche Grammatik für höhere Schulen, there appear to be five main versions:
[3.213] Starting around 1913 Mensing adds Satzbilder to analyse individual sentences with a graphical display. An example from the Vorstufe from 1914 in shown in Figure 3.90. Mensing explains that the words of the sentence are part of the graph, but that they can be replaced by functional abbreviations. Only slight cosmetic changes are made in the various later revision, as shown in Figure 3.91 from 1927 and Figure 3.92 from 1955.
(3.62) | a. | Die hellen Sternen zeigen dem einsamen Wanderer den Weg. |
b. | Der kleine Sohn des Bauerns brachte den durstigen Wanderer schnell einen Trunk kühlen Wassers. |
[3.214] In the 1954 revision of Mensing’s grammar by Petersen there is a graphical display added for the structure of complex sentences consisting of multiple clauses (Petersen 1955: 133-134). An example is shown in Figure 3.93. Main clauses are indicated by a low horizontal line and subordinate clauses by high horizontal lines. Start and end of a clause are shown with the small vertical lines, and double vertical lines indicate that the clause is separated into two parts by another clause. Multiply-embedded subordinate clauses are not specifically indicated in this approach.
[3.215] Johann Eberhard (born 1838) was a school director in Sigmaringen. Not much is available about his biography. He retired as director in 1903 and apparantly moved to Trier, see page 30 of the school report of 1904 https://nbn-resolving.org/urn:nbn:de:hbz:061:1-882473, accessed 10 January 2025. After his retirement he wrote a small article about the graphical analysis of sentences (Eberhard 1905), mostly as a reaction to the proposals of Stamm (see Section 3.3.32). Eberhard criticises Stamm for the complexity of his analyses, which are not suitable for use in school. However, Eberhard is very positive about the idea of using graphical displays for teaching grammar.
Das die graphische Darstellung ein erwünschtes Mittel ist, um Übersicht über ein verwickeltes Satzbilde zu erzielen, unterliegt keinem Zweifel. Nach meinen Erfahrungen bringen alle Schüler, auch die Minderbefähigten, der bezeichneten Darstellung ein großes Interesse entgegen – gewiss ein wichtiges Moment bei dem trockenen Lehrstoff! Die graphische Darstellung solcher Sätze, die in der Schule zergeliedert sind, ist deshalb auch eine viel begehrte Hausaufgabe. (Eberhard 1905: 284)
(‘There is no doubt that graphic representation is a desirable means of gaining an overview of a complicated sentence structure. In my experience, all students, even those with less ability, show great interest in the representation shown - certainly an important aspect given the dry subject matter! The graphic representation of such sentences, which are broken down in school, is therefore also a much sought-after homework assignment.’)
[3.216] Eberhard’s proposal for the graphical analysis of sentences is illustrated in Figure 3.94 for the example sentence in (3.63). Unfortunately, Eberhard still uses arbitrary numerals to indicate syntactic functions (Eberhard 1905: 283), but there are not very many of those, and they are used both for phrases and for clauses.
[3.217] His approach is quite close to a constituent analysis, as words are hierarchically grouped by the horizontal lines. There are three large groups, which are the main clause and two subordinate clauses, although there is no graphical indication of the fact that the third clause is recursively subordinate to the second. Inside each clause the various phrases are grouped. Eberhard refrains from adding more than two hierarchical levels (e.g. fremden might have been analysed as a third level), but this restraint is probably didactically intended. There are a few slight inconsistencies, but those seem mostly to be a problem with the printing (e.g. sich should not be part of the argument in the main clause, there is a line missing below 5c in the third clause, and the line at Kassenanweisung should also include the article).
(3.63) | Ehrfurchtsvoll verneigte er sich vor dem fremden Herrn, der diesen Augenblick benutzte, um still eine Kassenanweisung auf den Tisch zu legen. (Eberhard 1905: 284, citing Wilhelm Petsch Der Kronprinz Friedrich Wilhelm in Karlsbad) |
[3.218] Josef Bartmann (1868-1947) was a teacher in Nixdorf (today Mikulášovicích in the Czech Republic). Some biographical information is available on Wikipedia: https://cs.wikipedia.org/wiki/Josef_Bartmann, accessed 6 January 2025. He wrote various textbooks, one for teachers in 1907: Sprachübungen für die Hand des Lehrers (Bartmann 1913, cited here in the second edition) and another one for students in 1908: Deutsches Sprachbuch für Bürgerschulen (Bartmann 1917, cited here in the sixth edition).
[3.219] For the analysis of sentences, Bartmann uses two different methods. For single-clause sentences he proposes the method of zergliedern ‘to dissect’ (Bartmann 1913: 8-11), citing Kern (see Section 3.3.24). Like Kern, Bartmann uses line diagrams headed by the verb and he also compares these analyses to trees. The first example even has the verb at the bottom, with the verb being compared to the roots of the tree. But, also following Kern, Bartmann immediately reversed the display der besseren Raumverteilung wegen ‘for the better use of space’ (Bartmann 1913: 9). In the Sprachbuch he starts with a few simple line diagrams on the first page of the book (Bartmann 1917: 1), only to return to it with a bit more detail towards the end (Bartmann 1917: 174-176). An example of Bartmann’s diagrams is shown in Figure 3.95 for the example in (3.64).
(3.64) | Die Sonne neigt sich zum Untergange und die Blumen schließen ihre Kelche. |
[3.220] For complex sentences with multiple clauses Bartmann uses Satzbilder with various abbreviations for each clause (Bartmann 1913: 172-176; Bartmann 1917: 106-109). An example of a complex Gliedersatz in the analysis of Bartmann is shown in Figure 3.96 for the example in (3.65). Bartmann’s abbreviations are an idiosyncratic mix of the many different proposals available in the literature:
(3.65) | erg Daß der Trunkenbold sein meist mühsam erworbenes Geld in der unsinnigsten Weise verpraßt; erg daß er nach und nach seine geistige und leibliche Gesundheit zerrüttet; erg daß sein gesellschaftliches Ansehen mehr und mehr schwindet; erg/2 daß er, beif2 was das Schlimmste von allem ist, erg/2 seine meist völlig schuldlose Familie in Not und Verderben stürzt: A das überlegt er nicht, B oder es kommt ihm erst zum Bewußtsein, zeit wenn es oft schon zu spät ist. |
[3.221] Gustav Schuberth (1850-1922) was a director of a school in Großenhain. In the 33th yearbook of his school he wrote a short paper (Schuberth 1908), exemplifying how Satzbilder can be used to analyse a text. He took the first few pages of Schiller’s Geschichte des dreißigjährigen Kriegs and analysed all sentences with a simplified one-lined Satzbild. As an example of his notational system, Schiller’s sentence in (3.66) is analysed by Schuberth as shown in Figure 3.97. I have added the abbreviations from the figure into the example below.
(3.66) | A/2 In keinem andere Falle, (v) als unter diesem, A/2 war es den schwächern Fürsten möglich, (g) die außerordentlichen Anstrengungen von ihren Ständen zu erzwingen, e wodurch sie der österreichischen Macht widerstanden; (A) in keinem andern Falle den Staaten möglich, (g) sich gegen einen gemeinschaftlichen Feind zu vereinigen. |
[3.222] Konrad Bessel Erman (1888-1915) died a young man in the First World War, shortly after he finished his dissertation in Bonn (Erman 1913a; the actual dissertation was only the first part, see Erman 1913b). A short obituary is given by Schulze (1917) and a biographical eulology is provided by his mourning father (Erman 1916).
[3.223] In his dissertation, called Beziehungen zwischen Stellung und Funktion der Nebensätze mehrfacher Unterordnung im Althochdeutschen, Erman investigates different kinds of subordination in Old High German texts. He uses a symbolic representation to characterise the structure of complex sentences (Erman 1913a: 4-5), but refrains from calling them Periodenbild or Satzbild, although he cites various authors that use Periodenbilder for similar purposes (viz. Lehmann 1856; Esser 1878; Toifel 1897; Toifel 1901).
[3.224] Erman’s symbolic representation is rather different from his predecessors. An example of his display is shown in Figure 3.98 for the Old High German sentence in (3.67 a). This sentence is from the 9th Century Tatian translation of the the Gospel of Luke. A word-by-word etymological equivalence into modern German is given in (3.67 b). The Old High German translation follows the Latin structure (3.67 c), in which the verb venit of the first subordinate clause (marked b in the graphic) is separated by the second subordinate clause (marked c in the graphic). All this is placed before the main clause (marked a in the graphic).
(3.67) | a. | b ouh after thiu theser thin sun, c ther thar fraz alla sina heht mit huorun, b quam, a arsluogi imo gifuotrit calb. (Erman 1913a: 39, citing Tatian 97,7,7) |
b. | b auch achter-dem dieser dein Sohn, c der dort fraß all seine Echt mit Huren, b kam, a erschlägst ihm gefuttertes Kalb. (Modern German etymological equivalence) |
|
c. | b sed postquam filius tuus hic c qui devoravit substantiam suam cum meretricibus b venit a occidisti illi vitulum saginatum. (Vulgate, Luke 15:30) |
|
d. | But when this son of yours came, who has devoured your property with prostitutes, you killed the fattened calf for him! (English Standard Version, Luke 15:30) |
[3.225] Wilhelm Pfannkuchen (born 1887) wrote a dissertation in Giessen with Otto Behagel (Pfannkuchen 1914). I have not been able to find any further reliable information about his biography. Pfannkuchen writes in his dissertation that he started teaching at the Ludwig-Georgs-Gymnasium in Darmstadt in 1914 (Pfannkuchen 1914: 67). The yearbook of that Gymnasium for 1915 notes that he is let go to become the director of a private school, see page 10 of https://nbn-resolving.org/urn:nbn:de:hbz:061:1-718811. Pfannkuchen appears to have edited some plays by Schiller in 1930. There is an entry for a Dr. phil. Wilhelm Pfannkuchen in the Addressbook of Gotha in 1949 http://www.adressbuecher.net/addressbook/entry/54747dd11e6272f5d1ca01dc. Pfannkuchen’s dissertation Periodenbau in Goethes und Schillers größeren Dichtungen is basically a long list of references to sentences in the works of Goethe and Schiller, organised by the structure of their subordination. For complex sentences, consisting of three or more clauses, he uses a graphical display to illustrate their structure (Pfannkuchen 1914: 39-63). He cites various previous authors who use a graphical display with similar goals (viz. Lehmann 1852; Diringer 1878; Gelbe 1880; Kern 1883b; Schulze 1883 as discussed here previously), but in the end Pfannkuchen’s graphics are a new combination of these widespread ideas.
[3.226] An example of Pfannkuchen’s graphical analysis is shown in Figure 3.99 for the Periode from Friedrich Schiller in (3.68). The top-down ordering indicates depth of embedding and the left-right ordering corresponds to the actual ordering of the clauses in the sentence. Dots are used when a clause is separated by another clause and the plus-symbols is used for coordination.
(3.68) | Kond Und hast du dem Befehle deines Gottes, Relſ. Der in des Feld dich rief, ..ſ. genug getan Ausſ. So wirst du deine Waffen von dir legen +Ausſ. Und wiederkehren zu dem sanfteren Geschlecht, Relſ. das du verleugnet hast, +Relſ. das nicht Berufen ist zum blut’gen Werk der Waffen. (Pfannkuchen 1914: 56, citing Schiller Jungfrau von Orleans) |
[3.227] Charlotte (Lotte) Müller (1893-1972) was a teacher and pedagogue at the university of Leipzig, propagating the didactic principles of Hugo Gaudig. After the division of Germany she became a school director in West Berlin.
[3.228] Her first published work is a little booklet from 1921 Vom Deutschunterricht in der Arbeitsschule (Müller 1922, cited here from the second edition). In this book she describes in detail how a teacher could approach the analysis of a sentence like (3.69) and might end up with a graphical analysis as shown in Figure 3.100. Rather curiously, this graphic seems to consists of two different parts, placed on top of each other. The two parts are headed by the subject noun Die Klasse ‘the class’ and predicate spielte ‘played’, respectively. However, from Müller’s description it becomes clear that these two parts are supposed to be placed side-by-side and it is probably simply due to the lack of space on the printed page that they are place on top of each other. Placing them side-by-side puts this line-based graphical display in the tradition of Utzinger (see Section 3.3.27) and Mensing (see Section 3.3.35).
(3.69) | Unter den schattigen Bäumen auf der Debrahofwiese spielte gestern die wilde Klasse des Herrn R. ganz toll ein lustiges Spiel, Sackhüpfen. |
[3.229] Marius Hendrikus Flothuis (1873-1950) was a teacher of German language and literature in Amsterdam and the father of the composer with exactly the same name. Registration from the Amsterdam archive: https://archief.amsterdam/indexen/deeds/98533401-2df7-56a3-e053-b784100ade19?person=98533401-2df8-56a3-e053-b784100ade19. He wrote a German grammar Einfache kurzgefaszte deutsche Syntax, which contains many examples of graphical analyses of complex sentences called Satzbilder (Flothuis 1930: 136-145). Throughout his book, Flothuis is exceptionally precise in referencing the works of other scholars, but the section about the Satzbilder does not make any reference to previous approaches. Possibly, he considered the graphical display to be his own invention.
[3.230] Flothuis’ Satzbild for the example in (3.70) is shown in Figure 3.101. He uses a music-score metaphor with the main clause on top and subordinated clauses positioned downwards. The left-right order follows the actual ordering of the clauses in the sentence. The graphic includes the syntactic function of each clause, not the actual content. Parts of clauses, that are separated by another clause, are connected by a small horizontal arrow.
(3.70) | kausaler N.-S. Da seine Wohnung nahe am Eschenheimer Tore lag, Haupt- so führte mich, temporaler N.-S. wenn ich ihn besucht hatte, Satz mein Weg gewöhnlich zur Stadt hinaus und adv.Best.z.d.Grundst. zu den Grundstücken, unterscheidender Rel.-S. welche mein Vater vor den Toren besaß. (Flothuis 1930: 136, citing Goethe Dichtung und Wahrheit) |
[3.231] Franz Wollmann (1871-1961) was director of the teacher’s seminary in Krems and later in Vienna (Austria). Basic biographical information is available on Wikipedia at https://de.wikipedia.org/wiki/Franz_Wollmann, accessed 11 February 2025. The list of publications on Wikipedia does not correspond to entries in various library catalogues that I checked. He co-authored various textbooks with Sprachübungen ‘language exercises’ for teaching German. This series of textbooks was initiated by Bernhard Merth (1864-1922). Some basic biographical information on Bernhard Merth is available online at https://www.biographien.ac.at/oebl/oebl_M/Merth_Bernhard_1864_1922.xml. There are many different editions with numerous different booklets for different grades in this series. The many editions are in need of a more detailed investigation (an early example without graphical grammar is Merth & Wollmann 1909).
[3.232] Towards the end of his active career, Wollmann wrote the Deutsche Sprachkunde auf sprachgeschichtlicher Grundlage (Wollmann 1935), which appeared mostly unchanged in various editions at least up to a seventh edition in 1963. In this book he applies various different methods for visual syntactic analysis. First, clause-internal structure is visualised using a dependency tree as shown in Figure 3.102 for the examples sentence (3.71 a).
[3.233] For complex multi-clausal sentences he uses a simplified variant of a Periodenbild, only using the capital letter H=Hauptsatz ‘main clause’ and the lower-case letter n=Nebensatz ‘subordinate clause, as shown in Figure 3.103 for the example sentence (3.71 b). Multiple subordination is indicated by lower placement. Non-finite clauses are placed inside round brackets (here s.b.=Satzbestimmung ’clausal modifier’). As an alternative visualisation, Wollmann also proposes to analyse multi-clause sentences using the dependency model, but now the individual nodes are complete clauses. An example of this approach is shown in Figure 3.104 for the example in (3.71 c).
(3.71) | a. | Die gute Mutter bringt dem noch im Bett liegenden kranken Kinde heißen Tee gegen den bößen Husten. |
b. | n Wenn wir sehen, n daß die Sonne wieder wärmer scheint, H gehen wir ins Freie (s.b.) um nach den ersten Frühlingsblumen zu suchen. | |
c. | Auch dann, wenn du ein Buch liest, das nicht deinen Verstand bilden will, unterbrich zuweilen das Vergnügen, das dir die Erzählung verschafft, so lange, bis du mit ein paar Worten ausdrücken kannst, was der Verfasser in einem ganzen Abschnitt gesagt hat. |
[3.234] Johannes Heemstra (1882-1961) was a teacher for German in Haarlem (Netherlands). It is probably this person: https://gw.geneanet.org/pveenboer?n=heemstra&oc=&p=johannes, but the birthdate is probably wrong there, cf. https://www.erfgoedleiden.nl/collecties/personen/zoek-op-personen/deeds/f91c6f4d-a72c-7bf0-c389-d331292bb661?person=c4f023b1-de49-408f-eb22-ed91108d4107, both pages accessed 8 January 2025. He studied in Leiden, defended a PhD-thesis about German participles in Groningen and later wrote various textbooks for Dutch students of German. After his retirement he wrote the Grundrisz der deutsch-niederländischen Satzlehre (Heemstra 1947). In this book he used a Satzbild as shown in Figure 3.105 for the example sentence in (3.72). This display is inspired by Kern (see Section 3.3.24), whose approach is discussed explicitly by Heemstra (Heemstra 1947: 77-78). For good measure, Heemstra also reproduces an example from Flothuis (see Section 3.3.42) as another possibility to illustrate the structure of a complex sentence (Heemstra 1947: 78-79).
(3.72) | NS 1 Als er nach Hause kam, HS fragte er seine Frau, NS 1 was der Arzt, NS 2 nach dem sie geschickt hatte, NS 3 weil das jüngste Kind plötzlich erkrankt war, zu seinem Befinden gesagt habe. |
[3.235] Robert Killinger (1925-2006, no further biographical information available) and Alfred Doppler (born 1921, professor for literature in Innsbruck, Austria), Some basic biographical information on Alfred Doppler is available on Wikipediat at https://de.wikipedia.org/wiki/Alfred_Doppler, accessed 11 February 2025. wrote a somewhat amusing German textbook intended to be less dry than a regular grammar. It is called Deutsch: richtig gesprochen, richtig geschrieben and was originally published in 1950 in Austria with the subtitle Ein unterhaltsames Buch zur Aneigung einer fehlerfreien Sprache, zur Vervollkommung vorhandenen Wissens und zum Nachschlagen in Zweifelsfällen ‘An entertaining book to acquire error-free language, to perfect existing knowledge and for reference in cases of doubt’. It is cited here from the third edition for a German audience (Killinger & Doppler 1955). In the discussion of Satzbau ‘syntax’ they use a verb-centric dependency tree, shown here in Figure 3.106 for the example sentence in (3.73). For this graphical analysis they cite Wollmann (see Section 3.3.43).
(3.73) | Ein guter Mensch in seinem dunklen Drange ist sich des rechten Weges wohl bewußt. |
[3.236] Killinger remains active as an author of textbooks for many decades, specifically with the series Sprachübungen. This series was originally established by Bernhard Merth (1864-1922) and was also temporarily co-authored by Wollmann (see Section 3.3.43). There are many different editions with numerous different booklets for different grades in this series (an early example without graphical grammar is Merth & Wollmann 1909). I have not been able to inspect all these different versions and editions. Notably, in a version for the sixth grade, Killinger (1975: 88) uses the music-score metaphor for the analysis of a complex multi-clause sentence (3.74) as shown in Figure 3.107
(3.74) | 1 Den Knecht legten sie gebunden auf die Erde, 2 steckten ihm ein Sperrholtz ins Maul und 3 schütteten ihm einen Melckkübel voll garstig Mistlachen-Wasser in Leib, 4 das nanten sie eine schwedischen Trunck, 5 der ihm aber gar nicht schmeckte, 6 sondern in seinem Gesicht sehr wunderliche Minen verursachte, 7 wodurch sie ihn zwungen, 8 eine Parthey anderwerts zuführen, 9 allda sie Menschen und viehe hinweg namen und 10 in unseren Hof brachten, 11 unter welchen mein Knän, meine Meuder, und unsere Ursele auch waren. (Killinger 1975: 88, citing Hans Jakob Christoffel von Grimmelshausen Der Abentheuerliche Simplicissimus Teutsch) |
[3.237] Barbara Langholf wrote a dissertation at the university of Hamburg in 1969. She appears not to have pursued an academtic career after her dissertation as I have not been able to locate any subsequent scholarly works by her. Her dissertation investigates syntactic structures in the 16th Century anonymous German translation of the French novel Amadis de Gaule. The syntactic analyses used by Langholf are a direct descendant of the Periodenbilder of the 19th Century, although by the time of her dissertation the syntactic theories of Chomsky and Tesnière already had become widespread in the linguistic community. Langholf does not seem to have been aware of (or not interested in) those theories.
[3.238] For her syntactic analyses, Langholf devised her own scheme, which she calls a Formel (1969: 10). She cites a 1955 edition of Menge’s Repetitorium (cf. Section 5.1.2) as inspiration for her formulas. Menge’s approach dates back to 1873, almost 100 years before Langholf’s dissertation. Langholf’s (and Menge’s) formulas include various notational aspects that can be traced back to Lehmann’s original approach from 1833 (see Section 3.2.8), including the colon for preposed clauses and brackets for internal clauses. Langholf uses forward slashes for postposed clauses, a proposal originally from Nägelsbach in 1848 (see Section 5.1.1). Similar to Lehmann, Langholf uses different kind of symbols for the depth of embedding:
[3.239] Langholf adds a functional category to each clause, written as a superscript abbreviation (e.g. rel=relative clause, obj=object clause, etc.). She also adds superscript numbers as labels to identify the clauses. This is particularly useful to link the analysis to the original sentence and to connect separated parts of the same clause, which often are positioned far apart (e.g. clause 4 in the example below). As an example of her approach, consider the example in (3.75) and the corresponding analysis in Figure 3.108. Langholf notes that this example is extraordinary complex. Such complex constructions typically appear when the sentence is not in the French original, but added by the anonymous German translator.
(3.75) | 1Awunsch Doch daß E.May. zuuor bey Königlichem glauben, 2(arel) der in allen dingen vor jederman gantz warhafftig vnd der tugend anhengig sein soll, 1Awunsch und bey Ritters Orden, 3(brel) welcher zu handthabung der trew vnd gerechtigkeit erfunden vnd eingesetzt, 1Awunsch bestettigen vnd zusagen, 4/cobj daß, 5(a’kond wo E.May. mit meinen gnedigsten Freuwlin Elisena so weit handlen, vnd sie dahin bewegen möchten, 6/1obj das jhr biß zu vollstreckung der Liebe fürschritten, 7/Irel (welches gleichwol ich nicht gedenck, angesehen jr hohe Keuscheit vnd fürsichtigkeit, 8/IIfin Nichts desto weniger aber, damit allem, 9[xrel] so begegnen möcht, 8IIfin genugsam, vnd bey der zeit fürkommen, vnd die Ehr, als das höchste Gut, in allweg allerseits erhalten werde, 10/IIIkaus Denn Menschliches, insonders Weibliches Geschlecht gar schmach, 11/ykons also daß mein gnedigstes Frewlin, 12<x’rel> die vielleicht E.May. gleichermassen hertzlichen liebet, 11ykons) sich etwas vnbedachts vbersehen wünde) 4cobj das denn zumal vad hernach, 13(b’rel)wo es die zeit vnnd not erfordert, 4cobj E.May. sie zu dero Ehegemahel auffnemmen wöllen. (Langholf 1969: 141-142, citing page 19-20 of the Feyerabend edition) |
[3.240] Trunk mentions other grammars, but not available online https://babel.hathitrust.org/cgi/pt?id=uiuo.ark:/13960/s2zqj4w2gwz&seq=60
[3.241] nothing yet in 1880 version of Sprachschule, only in Binstorfer revision?:
[3.242] Also: Bernhard Merth (1864-1922, Sprachübungen, Sprachbuch, Sprachunterricht) => Wollmann, Killinger, Sernko as co-authors/revisers. no apparent graphics.
[3.243] Ernst Götzinger (1837-1896), son of Maximilian Götzinger (see Section 3.3.1), was a teacher in St. Gallen (Switzerland). He wrote extensively about local history, German literature and he reissued his father’s books after his death (more biographical details in Dierauer 1897). He also wrote a grammar, called Deutsche Grammatik in genetischer Darstelllung (Götzinger 1880). Concerning graphical display, Götzinger (1880: 151) criticised the Periodenbilder of Lehmann (see Section 3.2.8) and proposes an alternative. Intriguingly, he does not mention his father’s graphical approach, which is highly similar to the proposals by Lehmann.
[3.244] Götzinger proposes a very simple model for sentence analysis, consisting only of a plus symbol and hierarchically ordered brackets (Götzinger 1880: 155-156). He uses the term algebraische Gleichungen ‘algebraic formulae’ for his approach. Although the method is described very clearly in words, as quoted below, Götzinger does not give any examples of how his grammatical algebra works in practice for more complex sentences. A few concrete examples of Götzinger’s approach are given much later by Dinkel (see Section 3.4.2).
Will man das Satzgefüge in schematischen Bildern darstellen, so wird man sich besser der Methode der algebraischen Gleichungen bedienen. Angenommen, S heiße Subject, SS Subjectsubstantiv, P Prädicat, PV Prädicatverb, so ist die Formel jedes Satzes: S+P oder SS+PV. Beide Satzglieder lassen sich nun erweitern; ist das Subject-substantiv durch ein Adjectiv (a) oder durch eine Apposition (A) oder durch einen Genitiv (G) oder durch einen indirecten Casus (IC) erweitert, so würden die Formeln entstehen: (a+SS)+PV; (SS+A)+PV; (SS+G)+PV oder (G+SS)+PV; (SS+IC)+PV. Wäre das Prädicatsverb durch einen Accusativ (Acc.) oder Genitiv (Gen.) oder Dativ (D) oder einen indirecten Casus (IC) oder durch ein Adverb erweitert, so würde sich der Buchstabe dafür mit PV zusammenklammern lassen müssen, z.B. SS+(PV+Gen). Jede neue Erweiterung beiordnender oder unterordnender Natur ließe sich dergestalt in die Grundgleichung einfügen, ebenso jeder Nebensatz, der entweder dem Subject oder dem Prädicat angehört; das wesentlichste Mittel dieser Satztypen ist immer die Klammer, vermittelst welcher die zu einer begrifflichen Einheit zusammentretenden Einzelglieder aneinandergefügt werden. (Götzinger 1880: 156)
(‘When one wants to display the sentence structure in a schematic image, then it is to be preferred to use the method of algebraic equations. Suppose S is called subject, SS subjective substantive, P predicate, PV predicative verb, then the formula for each sentence is: S+P or SS+PV. Both these constituents can be expanded; when the subjective substantive is expanded with an adjective (a) or with an apposition (A) or with a genitive (G) or with an indirect case (IC), then the following formulae would arise: (a+SS)+PV; (SS+A)+PV; (SS+G)+PV or (G+SS)+PV; (SS+IC)+PV. When the predicative verb is expanded with an accusative (Acc.) or a genitive (Gen.) or dative (D) or an indirect case (IC) or with an adverb, than the letter for this expansion would be encapsulated together with the PV, e.g. SS+(PV+Gen). Likewise, each new expansion of a coordinating or subordinating nature can be inserted into the basic formula, and likewise each subordinate clause, which belong either to the subject or the predicate. The crucial method for these [more complex, MC] sentence types is the bracket, by which the elements, which are to be assembled into a conceptual unit, are joined together.’)
[3.245] Georg Dinkel (dates unknown) from Nürnberg wrote a small article in the Bayerische Lehrerzeitung about his Plan einer einfachen Grammatik des Deutschen auf historischer Grundlage ‘plan for a simple German grammar on a historical basis’ (Dinkel 1895). In this article he proposes a Satzbild that is based on the proposals from Ernst Götzinger (see Section 3.4.1), who is explicitly cited by Dinkel. Götzinger himself did not give any concrete examples of his approach, so the few examples in Dinkel’s article are the only available examples of this kind of analysis. Of particular interest is the principle to use nested brackets to indicate hierarchical structure. For example, the sentence in (3.76) is analysed with the Satzbild in Figure 3.109
(3.76) | Ew Ein munterer Sennbub 3. aus einem Dorfe 3. in der Schweiz P stieg 2. eines Tages 3. auf einem steilen Gebirgspfade 4. ins Thal herab. |
[3.246] Adolf Stöhr (1855–1921) was professor for philosophy in Vienna (biographical details in Angetter 2009). He wrote a rather curious book called Algebra der Grammatik (Stöhr 1898), which was assessed as “rather fantastic” by Jespersen (1937: 83). Although idiosyncratic, Stöhr’s book is not as crazy as Jespersen makes it sound. Stöhr assigns codes to linguistic elements, which are subsequently combined into larger expressions. However, his primary goal does not appear to be any detailed description of linguistics structures, but rather an attempt to try and “calculate” with these larger expressions, somewhat akin to what today would be called deriving semantic inferences.
[3.247] Stöhr (1898: 6-11) differentiates between elements a (=Ausdruck, today we would say “lexical morpheme”) and d (=Derivation, today we would say “grammatical morpheme”). Different morphemes are distinguished by a numeric identifier that is written above the letter. Stöhr goes into great detail for the grammatical morphemes, defining many different kinds of d elements. Although he explicitly states not to pursue completeness (nicht ein erschöpfendes Verzeichnis, Stöhr 1898: 15), I have counted more than 200 different defined codes in his book, classified into 15 different classes indicated by different letters. Just scrolling through his book, I found 46 different grammatical elements designated as i, 47 elements called m, 62 times o, 6 times v, 9 times t, 5 times e, 2 times p, 3 times n, 4 times b, 2 times h, 4 times l, 4 times f, 11 times q, 2 times g and 2 times s. It is beyond the scope of this book to go into more detail here.
[3.248] Elements can be combined by writing them side-by-side, called Ableitung, which is typically used for examples of composition and derivation (in modern terminology). This Ableitung is opposed to Apposition (Stöhr 1898: 51ff.) for syntactic combination, written with a star-symbol in between the elements. Stöhr notes that such Apposition is hierarchically organised and he proposes to use brackets to indicate this hierarchy. He exemplifies this with the example (3.77) as graphically represented in Figure 3.110.
Das geeignetste Zeichen für die Apposition eines selbst wiederum durch Apposition gebildeten Ausdruckes zu anderen Ausdrücken, dürften zwei Klammern sein, zwischen welche man den zusammengesetzten Ausdruck schreibt. (Stöhr 1898: 87)
(‘The most suitable sign for the Apposition of an expression, which itself is formed by Apposition to other expressions, is probably two brackets, between which the compound expression is written.’)
(3.77) | a1 ein a2 Mann, o6 in a1 einem a3 langen a4 Mantel, o4 mit l4 dem a5 Hut o44 auf l4 dem a6 Kopfe |
[3.249] uses SVOPA abbreviations
[3.250] https://archive.org/details/grammatikderrom00meyegoog/page/799/mode/2up
[3.251] Friedrich Feigl, inspired by Meyer-Lübke
[3.252] https://opac.ku.de/s/uei/de/2/-1+%3D+%22Friedrich+feigl%22+filter%281280%3D%22%3C%3D1908%22%29
[3.253] Erich Drach (1885-1935) was a professor for speech science in Berlin (some basic biographical information is available in Winkler 1959). His posthumously published book Grundgedanken der deutschen Satzlehre (Drach 1937) is regularly cited as the foundation of the contemporary Felderanalyse as used widely in German syntax. However, while Drach indeed uses the term Feld ‘field’, his templatic analysis of the German clause is far from original nor innovative (see Figure 3.111). First, the idea to name positions in relation to the verb-position in a German sentence (Vorfeld, Nachfeld) was already used by Lehmann in 1833 (see Section 3.2.8). Further, Drach does not distinguish between what today would be called a Mittelfeld and a Nachfeld and simply collapses everything in one position called Nachfeld. Additionally, Drach’s analysis omits the crucial idea of a fixed position for non-finite part of the verbal complex, which today is often referred to as a rechte Satzklammer. This structural position was, for example, already identified by Lonkay in 1855 (see Section 5.5.1). Drach instead talks about the end of his Nachfeld as the Zielpol ‘goal pole’, but this is not a syntactic position in Drach’s analysis. The Zielpol is simply the focus of the sentence, which could just as well be an object or an adverb.
[3.254] Not widely discussed, but much more interesting, is Drach’s discussion of Umklammerung ‘bracketing’. He argues that one of the major challanges for stilistically well-build sentences in German is the syntactic tendency to insert constituents inside other constituents (rephrasing Drach’s approach in contemporary terminology here). This leads to bracketing because related parts of the utterance are separated by internal subordination. Such internal subordination is used very frequently in German and can lead to unmögliches Deutsch ‘impossible German’ (Drach 1937: 46). With this slightly ironic formulation Drach does not argue that the resulting structures are syntactically ill-formed, but that it is stylistically problematic.
[3.255] Drach uses different kinds of brackets to depict the levels of subordination, as illustrated in (3.78) reproduced from Drach (1937: 46, similar examples on page 39 and 49). The brackets clearly enclose parts that today would be called a constituent. The only difference to a modern constituency analysis is that Drach does not label these constituents. Other than that, Drach’s Umklammerung is completely analogous to a constituent tree, as shown in Figure 3.112.
(3.78) | Er { ging, [ ohne ( auf die < sich bei ihr wiederholenden > Klagen ) einzugehen ] aus dem Zimmer }. |
[3.256] Johann Leo Weisgerber (1899-1985) was a professor for comparative linguistics in Rostock, Marburg and Bonn. An extensive discussion of his life and work is provided by Roth (2004). In his syntactic analyses, Weisgerber was strongly influenced by the posthumously published work of Drach (see Section 3.5.1). Already in Die volkhaften Kräfte der Muttersprache (Weisgerber 1939) he cites Drach’s Grundplan ‘outline’ (see Figure 3.111) and he re-iterates the importance of Drach’s notion of Umklammerung ‘bracketing’ for the syntactic stucture of German (Weisgerber 1939: 54-58). After the war, Weisgerber published his 4-volume opus magnum, called Von den Kräften der deutschen Sprache, in which he again discusses the proposals from Drach in volume 3: Die Muttersprache im Aufbau unserer Kultur (Weisgerber 1950: Grundplan 120-121, Umklammerung 188-193). In this discussion Weisgerber includes various different visualisation for the hierarchical structure of the Umklammerung.
[3.257] The first visualisation is used for the example (3.79 a), which Weisgerber already used in his earlier work (Weisgerber 1939: 58). However, he now adds the hierarchical visualisation as shown here in Figure 3.113. This sentence is an example of a single hierarchical bracketing, but Weisgerber notes that it is perfectly possible, though slightly cumbersome, to have multiple such complex brackets in a single sentence. As an example he uses the sentence (3.79 b), visualised as Figure 3.114. This analysis uses Drach’s Grundplan (consisting of Vorfeld-Mitte-Nachfeld), but in a rather different way from all other examples from both Drach and Weisgerber. Normally only the verb fand ‘found’ would be placed in the middle, the rest would be placed in the Nachfeld. Also note that the predicative bracket fand…statt from the verb stattfinden ‘to take place’ is not considered as a bracket. This kind of “particle-verb” bracket is very common in German, and Weisgerber includes such structures as a bracket in the other examples shown here (3.79 a,c). For the more complex example in (3.79 b) he is clearly stimified by Drach’s simplistic Grundplan, which does not include a separate position for the verb particle. Such a position is commonly added in contemporary analyses of German (e.g. as a rechte Satzklammer). Finally, Weisgerber presents yet another example of a single hierarchically complex bracketing in (3.79 c) with a visually pleasing diagram as shown in Figure 3.115. In this illustration he tries to emphasize the ordering and the hierarchichal depth of the bracketing.
[3.258] In the second, much expanded, edition of this work, Weisgerber repeats the same illustrations (Weisgerber 1954: 190-192). Furthermore, he adds another approach for the visualisation of hierarchical structure by using a “chinese boxes” diagram with hierarchically enclosed rectangles (Weisgerber 1954: 252). However, this illustration is copied verbatim from the work of his student Gipper (see Section 3.5.3) and will be discussed there.
(3.79) | a. | Sie zogen [ in die [ reich [ mit [ bunten ] Fahnen ] geschmückte ] Stadt ] ein. |
b. | Vorfeld Am [ dem [ seiner [ unerwarteten ] Rückkehr ] folgenden ] Abend Mitte fand [ in dem [ lange [ von seinen [ rechtmäßigen ] Bewohnern ] verlassenen ] Schloß ] Nachfeld ein [ mit [ aller [ aus [ früheren ] Zeiten ] gewohnten ] Pracht gefeiertes ] Fest statt. | |
c. | Sie kamen [ mit [ einer [ infolge der [ plötzlichen [ in der [ vorangegangenen ] Nacht ] eingebrochenen ] Kälte ] ungewöhnlich großen ] Verspätung ] an. |
[3.259] Helmut Gipper (1919-2005) was a professor for general linguistics in Münster and a student of Weisgerber (see Section 3.5.2). The best biographical summary for Gipper that I could find is on the German Wikipedia at https://de.wikipedia.org/wiki/Helmut_Gipper. He finished his dissertation Sprachliche und geistige Metamorphosen bei Gedicht-Übersetzungen in 1950, but it was only published in 1966 in einen fast unveränderten Druck ‘an almost unchanged edition’ (Gipper 1966: 8). In the preparation of his dissertation, Gipper had access to a pre-print version of Weisgerber’s (1950) Von den Kräften der deutschen Sprache and he cites Weisgerber’s example (3.79 a) as an example of hierarchical embedding.
[3.260] Gipper develops his own visualisation using a “Chinese boxes” approach to show the hierarchical structure in einer besonders anschaulichen Form ‘in a particularly vivid manner’ (Gipper 1966: 118). As shown in Figure 3.116, he compares the central embedding of German with the right-modifying structure of the French translation. This visualisation is quoted verbatim by Weisgerber in the second edition of Von den Kräften der deutschen Sprache (Weisgerber 1954: 252). Gipper does not cite Pike (see Section 4.7.6), Diderichsen (see Section 5.2.7) or Mikuš (see Section 4.7.8), who also use this visualisation in the same period in which Gipper wrote his dissertation. Apparently, this visual idea became en vogue in the 1940s.
[3.261] Hanz Glinz (1913-2008) was professor for German philology in Aachen. A biographical summary and an in-depth discussion of his work is provided by Zlobinska-Görtz (2018). His Habilitationsschrift was accepted in Zürich in 1948 and published as Die innere Form des Deutschen. Eine neue deutsche Grammatik (Glinz 1952). This book was widely discussed in Germany and is generally considered to be an early example of a structuralist grammar for German. Glinz consistently uses Proben, i.e. syntactic constituency tests, to analyse the grammatical structure of German.
[3.262] Glinz’ book contains two large examples in which he uses a literary text to illustrate graphically how his analysis works in practice. The first example uses a paragraph from Goethe’s Wilhelm Meisters theatralische Sendung and a paragraph from Heinrich Keller’s Der grüne Heinrich (Glinz 1952: 71-72). The full analysis takes up ten whole pages (Glinz 1952: 316-325). The last sentence from this exemplary analysis (3.80) is reproduced here in Figure 3.117. This visualisation places the words from top to bottom with thick black indicators for the verbal elements of the sentence. Glinz very clearly includes both the “finite” second position and the sentence-final “infinite” position in his analysis of the German sentence. The arrows indicate the governer of the case assignment. In the complex noun phrases the horizontal lines indicate the hierarchical structure. Subordinate clauses (not shown in the figure here) are not visually distinguished in this large example.
(3.80) | und die kühle erfrischende Luft atmend schlief ich sozusagen an der Brust der gewaltigen Natur ein. (Glinz 1952: 325, citing Heinrich Keller) |
[3.263] Glinz’ second large example is introduced as a massive 10-page fold-out sheet inserted between pages 472 and 473. For this exemplary analysis Glinz uses a paragraph from a letter from Goethe, sent on the 12th of november 1779 and published in the collection Briefen aus der Schweiz. The graphical analysis of the first sentence of this letter (3.81 b) is shown here in Figure 3.118. The linked high-resolution image shows a panorama of the complete fold-out sheet. In this visualisation the words are ordered horizontally and there are lines added on top of the whole example. These lines indicates the depth of clausal embedding: deeper embedding is indicated by lines that are higher positioned.
(3.81) | Es war ein seltsamer Anblick, wenn man einen Moment seine Aufmerksamkeit vom Wege ab und auf sich selbst und die Gesellschaft wendet. (Glinz 1952: 472, citing Goethe) |
[3.264] https://ids-pub.bsz-bw.de/frontdoor/deliver/index/docId/8374/file/Erben_Abriss_der_deutschen_Grammatik_1958.pdf
[3.265] p.188: citing Tesniere (Tesnière 1953), in the 1964 edition p. 266 adds a quib about the simplicity of the american constituent analys is https://doi.org/10.1515/9783112763575
[3.266] vorfeld/nachfeld
[3.267] Very close to Drach (Esser 1961)
[3.268] https://ids-pub.bsz-bw.de/frontdoor/deliver/index/docId/595/file/Engel_Regeln_zur_Wortstellung_1970.pdf
[3.269] https://d-nb.info/112594563X/34 http://nbn-resolving.de/urn:nbn:de:bsz:mh39-5959
[3.270] https://www.persee.fr/doc/drlav_0754-9296_1980_num_22_1_957
[3.271] sentence template p. 63-64
[3.272] Erika Essen (1914-1986) bio (Müller 1996)
[3.273] (Essen 1956; Essen 1958) very nice curved lines!
[3.274] https://www.lagis-hessen.de/de/subjects/print/sn/bio/id/18993
[3.275] (Grosse 1960) indented subordinates clauses (Grosse 1966) many modern approaches
[3.276] (Rychener 1960)
[3.277] (Rychener 1982) Many different ideas by Rychener: underlining of clause-internal structure (Rychener 1982: 150-151), Becker-style tables (Rychener 1982: 178), tab-inserts (Rychener 1982: 176), Vertical displacement of clause-combinations like Flothuis (Rychener 1982: 189)
[3.278] (Meinel 1961: 21) lines with different height for subordination
[3.279] Satzbaukasten: (Ulshöfer 1967: 133-138)
[3.280] Satzbilder with lines in (Lobentanzer 1986: 60, 72-73)
[4.1] “Normal schools” < école normale, education for teachers!
[4.2] Greene (1836) A practical grammar of the English language “diagram” of tenses, not syntactic structure https://de.wikipedia.org/wiki/Roscoe_Greene https://google.com/books?id=PI4VAQAAMAAJ
[4.3] grammaire generale (Sicard 1798) numbers p. 29 “chiffre”, but only one example?
[4.4] Cours d’instruction (Sicard 1799) numbers? p 64-71, 101, 112, 178, 332-333. graphics: 89
The method of illustrating the principles of syntax employed in the following treatise, is one which has been, for a number of years, advantageously used in the instruction of the deaf and dumb. […] The first notions of the writer regarding it, were obtained while he was engaged as an instructor in the American Asylum, at Hartford, in the years 1831 and 1832. The method, which it at present exhibits, has been, however, the result of a more recent labor, during the writer’s residence, in a corresponding capacitiy, at the New York Institution for the Instruction of the Deaf and Dumb. (Barnard 1836: iii)
[4.5] Analytic Grammar; With Symbolic Illustration. New York: E. French. https://catalog.hathitrust.org/Record/008991837 No trees, but “symbolic sentences” in which categories are marked. Some ideas of levels http://www.polysyllabic.com/?q=node/336. Developed at school for the deaf in Hartford! based on l’epee/sicard “. This development seems to be linked to deaf-education, see (Wing 1887: 86)
[4.6] Francis A. March, A Parser and Analyzer for Beginners, with Diagrams and Suggestive Pictures (1869). http://www.polysyllabic.com/?q=olddiagrams/clarkderiv thanks to: Richard Salter Storrs (deaf education in hartford, mentioned in preface of March 1869), see (Williams & Harding 1885: 102)
[4.7] (Storrs 1880a; Storrs 1880b; Storrs 1881)
[4.8] (Wing 1885) further development of storrs/peet symbols for the deaf (which were also the basis for March 1869)
[4.9] Note tables in (Becker 1830: 253-261)
[4.10] “construction” in table-like manner like Becker 1830 (James 1847: 133-134)
[4.11] Morell (Morell 1857[1868]; Morell 1852) is based on Becker 1829! But simplifies into tables, and removes ordering. Tables become widespread in GB
[4.12] table p24-25, complex sentence p41
[4.13] William Swinton (1833-1892) in USA (Professor of the English language in the University of California??? Preface written in New York), but no diagram - more likely britisch tradition
[4.14] tabular based on Becker (Swinton 1872: 154)
[4.15] Swinton (1872: iii-iv) “The introduction, some thirty years ago, of the method of Sentential Analysis, devised by the German philologist Becker, and adapted to American school use in the meritorious works of Professor Greene and others, marks the only considerable innovation, in this country, on the Murray system.”
[4.16] He writes many versions of grammars and textbooks, but never uses diagramming. He cites many GB-grammars (p. viii), so this work seems to belong more in that tradition
[4.17] Also USA, but only using tables (Morris 1877: 100-111)
[4.18] first edition (Mason 1858) no tables 25th edition (Mason 1881: 183-185) with tables (fold-out appendix after p268)
[4.19] long-form analysis like becker (Fitch 1881: 268-269)
[4.20] graphical display “mapping” (Meiklejohn 1882: 20, 22, 147, 168-169)
[4.21] “mapping out” = diagramming? (Meiklejohn 1886: 92, 97-98) “continuous analysis = Becker-style? (Meiklejohn 1886: 99) tables p.98
[4.22] table (Cooper & Sonnenschein 1891: 9) indentation (Cooper & Sonnenschein 1891: 25)
[4.23] https://catalog.hathitrust.org/Record/006151225 table analysis p212, large example p220
[4.24] many different books and version, first collaboration from 1873
[4.25] tables (Nesfield 1898: 105-117)
[4.26] Charles Talbut Onions (1873-1965) Many editions, at least untill 1932
[4.27] tables and indented complex sentences (Onions 1904) sentence picture p.25
[4.28] (Wilbur & Livingston 1814) pictures for word classes. the “chart” mentioned could be this one: https://lccn.loc.gov/2020768232 (not digital)
[4.29] (Wilbur 1821) The grammatical key same picture for word classes, but apparently not used to draw pictures of sentences. In sentence analysis the numbers are used for parsing word classes. symbols and numbers are changed from the 1814 version
[4.30] word “chart” replaced by “atlas” in furthermore exactly the same sentences.
[4.31] “Construing consists in dividing a sentence into sections or groups, ascertaining their true constructive relation, learning their exact significant characters, and referring the inferior sections to their respective superiors.” (p. xii)
[4.32] (West 2000: 91-109)
[4.33] no graphics, but very concise description of dependency structure
In analyzing a proposition, it is first to be divided into its logical subject and predicate.
If the logical subject consists of more than one word, its grammatical subject should be pointed out, and distinguished as simple or compound.
When the grammatical subject is determined, the words which modify or limit it should next be specified, and then the words which modify them, and so on, until the logical subject is exhausted.
In analyzing the logical predicate, the grammatical predicate should first be mentioned, then the words which modify or limit it, and their modifiers, until the logical predicate is exhausted. [Andrews & S. (1836): 254-255
(4.1) | Mithridātes, duārum et viginti gentium rex, totǐdem linguis jura dixit |
The logical subject is Mithridātes duārum et viginti gentium rex.
The logical predicate is totǐdem linguis jura dixit.
The grammatical subject is Mithridātes: this is modified by rex. Rex is limited by gentium, which is itself limited by duārum and viginti. Et connects duārum and viginti.
The grammatical predicate is dixit, which is limited by jura and linguis, and the latter by totǐdem. (Andrews & S. 1836: 255-256)
[4.34] parsing inspired by Brown???
[4.35] (Barrett 1837) system of parsing (introduced to the learner in the preface vii-viii), key to analysis p. 29, parsing explained p31. Text of later image on p. 33.
[4.36] “scanning” on page 113.
[4.37] slightly different title, but probably just a new edition(Barrett 1845) insert between p.18 and 19. Parsing of the examples by letters, p.18-31
[4.38] (Barrett 1857), same graphic plate beween p18 and p19. Parsing exemplified p66-75
[4.39] new plates: at start, before p85 in latin
[4.40] note morpheme-separation by dashes in Latin p.125-131
[4.41] new book (Barrett 1859) proposing 21 different relations (p.28), the numbers are written to each word when parsing a sentence
[4.42] “chains” on p. 50, 52, 140 (Peirce 1839)
[4.43] “as the dependence of words in a sentence is more fully illustrated by the following figurative exemplifications, the pupil should study them till he shall be able, without difficulty, to apply to them any simple sentences that he shall find.” (Peirce 1839: 47)
[4.44] “immediate dependence” vs. “intermediate connection” (p. 141)
[4.45] numbered example sentences p. 233-241, there appears to be some implicit hierarchical ordering, but no clear explanation. Maybe numbering like Brown?
[4.46] Samuel Stillman Greene (1810-1883)
[4.47] cited here in printing from 1849. graphics “formula for a sentence” (Greene 1849: 70, 82, 111, 120, 167, 184) meaning of digits is unclear
[4.48] Later different systems, e.g.
[4.49] “Monology” inspired by Brown! Numbering to indicate constituents (“monos”) (Hall 1850: 197-305). Also bracketing with square and round brackets.
[4.50] identical in 1849 version. bad scan at https://archive.org/details/encyclopediaofen00hall/page/296/mode/2up?q=monology
[4.51] Not the same person as William D. Hall from 1898 diagrams.
[4.52] coincidence “monème” by Henri Frei La grammaire des fautes (1929) ???
[4.53] Stephen W. Clark (1810-1901) was a school principal of East Bloomfield Academy and wrote the influential book The science of the English language. A practical grammar: in which words, phrases, and sentences are classified according to their offices, and their various relations to one another. Illustrated by a complete system of diagrams. (Clark 1847) He produced revisions of this book and wrote variants for different readers up to the 1870s [e.g Clark (1870), see (Mazziotta 2016: 303-305 for the little that is known about Clark). A detailed investigation of the work of Clark is available in Mazziotta (2016; 2020b; 2020a).
[4.54] It remains unclear whether Clark has had any sources of inspiration for his graphical display. In the preface he writes that he has been testing his method at least since 1840. The “chains” of Pierce (1839)
for the convenience of teachers a manuscript grammar was prepared, which embodied the principles of the science and the Author’s mode of presenting it. These principles and this method have been properly tested by numerous and advanced classes during the seven years last past. (Clark 1847: iii)
[4.55] Term office ~ syntactic function?
[4.56] He uses the term diagram
[4.57] adjuncts are hanging, but also enclosed as boxes on p17!
[4.58] (Clark 1847) very interesting frontispiece in the first edition! Is later sold separately as a poster for schools
[4.59] graphical analysis (Holbrook 1859: 174), citing “Brown’s rules of syntax” (Holbrook 1859: 165), probably Goold Brown. He also cites “Clark’s grammar, revised edition”
[4.60] complete grammar in 1873, same diagrams (Holbrook 1873: 145-151)
[4.61] nicer scan here (later version, same examples): https://archive.org/details/newenglishgramma00holb/page/224/
[4.62] 1860 (Chandler 1860) version has no diagrams, added in revised edition 1861 (Chandler 1861)
[4.63] “models for black board exercises” (p151-160) very clear inspiration for later Reed/Kellog (Reed/Kellog say in introduction that their system arose in 1868, so clearly later than chandler)
[4.64] Richard Brantley York (1805-1891) Brantley York was an educator, author, and Methodist clergyman in North Carolina. He organized Union Institute Academy at Brown’s Schoolhouse in Randolph Co., N.C. in 1839, which would evolve into Normal College, Trinity College, and later Duke University.
[4.65] not yet found: (1854) An illustrative and constructive grammar of the English language
[4.66] first edition of Analytical grammar from 1860
[4.67] https://en.wikipedia.org/wiki/Brantley_York
[4.68] 1862: 38, 39 (note brackets and numbers like Brown???)
[4.69] https://archive.org/details/analyticalillust00york
[4.70] p.21 !!!
[4.71] https://archive.org/details/yorksenglishgra00york/page/20/mode/2up?q=diagram
[4.72] https://catalog.hathitrust.org/Record/011199495
[4.73] diagram, tree metaphor in pictures
[4.74] https://books.google.de/books?id=BdIOAQAAMAAJ&newbks=1&newbks_redir=0&printsec=frontcover&hl=de#v=onepage&q=diagram&f=false
[4.75] following clark (1847)
[4.76] https://google.com/books?id=dmwVAQAAMAAJ
[4.77] no graphics in (Burtt 1859)
[4.78] (Burtt 1868) cites chandler in introduction (though not about graphics). completely revised and new diagrams (Burtt 1868: 263-279), improved from Chandler
[4.79] same in (Burtt 1873: 126-136), variant of Chandler
[4.80] William Henry Parker, original not yet found (Brittain 1973: 53)
[4.81] original not yet found (Brittain 1973: 67)
[4.82] https://bsky.app/profile/coffeeanddonatus.bsky.social/post/3kifj7o36bu2i https://nationallibraryofeducation.on.worldcat.org/search/detail/974041820?queryString=chamberlin%20natural%20system%20grammar&clusterResults=true&groupVariantRecords=false
[4.83] probably G.E. Lighthall (1872) Introduction to analysis and parsing with bubbles, original not yet found (Brittain 1973: 73)
[4.84] (Lighthall 1874), clark but without bubbles, just underline
[4.85] originally A practical grammar from 1845 without diagrams, see https://google.com/books?id=-khKAAAAIAAJ for 1846 revision (Butler 1846) note: “The rules for the Analysis of Sentences have been taken, with some changes, from Andrews and Stoddard’s Latin Grammar” (preface p.3) examples like Andrews/Stoddard on p220-221
[4.86] revised edition with slightly different name, with diagramming like Holbrook (Butler 1874: 178-187). Also interesting footnote in preface p5.
[4.87] Revised edition with original name A practical grammar in 1879 with diagrams (Butler 1879: 168-177)
[4.88] https://www.letsdiagram.com https://natureofwriting.com/course/diagramming-sentences
[4.89] Thomas Wadleigh Harvey (1821-1892)
[4.90] practical grammar (Harvey 1868) no graphics. revised practical grammar (Harvey 1878: 266-272) with diagram like reed/kellog
[4.91] ??? not found: Harvey (1900) A new English grammar for schools (Brittain 1973: 178). Note that there is a 1900 revision of the elementary grammar by Louise Connolly https://lccn.loc.gov/00002828
[4.92] original elementary grammar (1869) https://lccn.loc.gov/11006142
[4.93] first attempt (Lyte 1879)
[4.94] expanded later in (Lyte 1899; Lyte 1898; Lyte 1886)
[4.95] William Francis Lewis Sanders (1849-1930)
[4.96] Born on Sept. 12, 1849, in Maxville, Ind., William Francis Lewis Sanders is the son of Jonas Bedford and Delphena Nevitt Sanders. He received the B.A. degree (1873) from Indiana University. On March, 10 1875, he married Fannie Taylor and they had five children: Pearl, Bertram, Emma Allison, Mabel Rosalie, and Earl Prentiss. Sanders held various teaching and administrative positions in Indiana public schools at Owensville, Bloomington Cambridge City, New Albany, and Connersville . He died on Oct. 8, 1930.
[4.97] 1879 book not yet found, see (Brittain 1973: 89)
[4.98] 1891 The English sentence diagrams included at (Sanders 1891: 68-end)
[4.99] first edition 1881, cited here revised edition from 1885. Diagramming (Eubank 1885: 113-200) “brace system”
[4.100] original not yet found (Brittain 1973: 102) only cover page here from 1888 edition: https://digital.klnpa.org/digital/collection/philips/id/8878/
[4.101] vertical analysis (Raub 1880a: 129-156) “written analysis”
[4.102] also (Raub 1880b; Raub 1885) both referring to the first book as the origin. 1885 seems most comprehensive summary
[4.103] (Adams 1882) based on Holbrook and Harvey. Says Holbrook approach is widespread in America ?!
[4.104] (Vaughn 1883: 82-end) following reed/kellog
[4.105] reed/kellog-style (Patterson 1884: 206-220)
[4.106] (Sornberger 1884: 80) vertical SP-diagram like Holbrook (author is from a “Normal school”)
[4.107] Frank Van Buren Irish (Irish 1884) like reed/kellog
[4.108] “sander’s system of diagrams’ preface p. iii, note that Harvey has many diagrams himself
[4.109] original not yet found (Brittain 1973: 118)
[4.110] Isaac Eldridge Wilson
word-class tree: p. 488, 494 (see Blanchard: Wood 2000: 138, originally 1847? https://catalog.mwa.org/vwebv/holdingsInfo?searchId=3288&recPointer=0&recCount=10) https://www.loc.gov/item/2018757044/
diagramming: (Wilson 1886: 544-554)
[4.111] Alfred H. Welsh (1850-1889)
[4.112] no graphical analysis:
[4.113] graphing:
[4.114] after death, edited by greenwood, based on welsh’ “lessons” 1887:
[4.115] diagramming (Rigdon 1887: 13-26, 58-68)
[4.116] Grammar of the English sentence (1890) diagramming like reed/kellog p.167
[4.117] https://books.google.de/books?id=dJAOAAAAYAAJ&printsec=frontcover&hl=de#v=onepage&q&f=false
[4.118] Harris R. Greene (Greene 1889: 36, 41, 49, etc.). combination of “reverse” tree and symbols below the worlds.
[4.119] first printing from 1888, cited here unchanged 1889 version
[4.120] earlier books do not seem to have graphics
[4.121] Judson Perry Welsh (1857-1934) Bloomsburg normal school president 1890-1906, https://library.bloomu.edu/pages/Archives/Presidents/Welsh.html
[4.122] thanks teacher Francis A. March at lafayette college in the preface (p. v). diagram (Welsh 1889: 29ff.)
[4.123] rather different kind of diagrams from the other traditions. Own invention? Grammar based on Goold Brown. (Kiddle 1889: 44-)
[4.124] word-analysis (phonetics, orthography, syllables), parsing (word classes, morphology), sentence-analysis (syntax, “scheme” p.304), scanning (prosody)
[4.125] “They [the authors] also claim, as a new and good feature, the Syntax Tree; which follows and, as an illustration, explains Sentence-analysis in Part Third.” (p. vi)
[4.126] tree-analysis p326-336. Note words ARE the branches!
[4.127] diagrams for english in first part called Inductive studies in English grammar (Harper & Burgess 1891: 42-43)
[4.128] Later separate book has similar diagrams (Harper & Burgess 1894: 59-60) in preface they cite grammars of Whitney, Meiklejohn, Salmon (Longman’s grammar???) and Welsh (which Welsh???)
[4.129] William Henry Maxwell (1852-1920) (Maxwell 1891)
[4.130] analysis simple sentences similar to reed/kellog but not identiacl 52-58, complex sentence completely different, using template-boxes among other methods: 280-297
[4.131] original not yet found (Brittain 1973: 151)
[4.132] (Robbins 1893) note: “revised edition”, but earlier version cannot be found. Vertical diagrams all throughout
[4.133] original not yet found (Brittain 1973: 156)
[4.134] Inductive Grammar. A Manual of Direction for the Study of English based on the Practical Grammar of Noble Butler by J. T. Gaines, O. B. Theiss
[4.135] Note very short mentioning of Butler with a single unexplained grammatical diagram on p43 of Principles in teaching (1891) https://catalog.hathitrust.org/Record/009568553
[4.136] diagrams (Buck 1894: 15-16, 111, 136)
[4.137] more extensive subordination diagrams (Buck 1900: 16-17, 117-118, 143)
[4.138] diagrams (Park 1894: 14, 16, etc.) cite Sanders for diagrams (p5)
[4.139] also (Park 1898: 107-109, 114-118)
[4.140] first edition from 1895, not yet found.
[4.141] Preface to the second edition (1896) says “a simple system of diagrams has been added” (p6). Cited here is the third edition (1897)
[4.142] The aim should be, not to diagram the sentence, but to analyze it. The diagram is only an aid to analysis. (p8)
[4.143] Diagramming (Hoenschel 1897a: 65-66), and throughout.
[4.144] Many more diagrams in the teacher’s companion (Hoenschel 1897b)
[4.145] diagrams (Hall 1898: 174-183) not the same Hall from “monology” in 1849
[4.146] diagrams (Bartlett 1899)
[4.147] David Salmon originally wrote School grammar (Salmon 1890), which was reissued as Longman’s school grammar (1891). This grammar does not contain any graphical analysis of sentences
[4.148] George J. Smith revised the book as Longman’s English grammar, though most of the text still is Salmon’s original. However, Smith (or somebody) added the graphical analysis (Salmon 1901: 266-271)
[4.149] In 1988 a completely different grammar appeared under the name Longman English grammar by L.G. Alexander. No graphical analysis.
[4.150] (Garrison 1901)
[4.151] refers to examples from English grammar for common schools, Robert C. Metcalf & Thomas Metcalf (1894) they give no diagrams https://archive.org/details/englishgrammarfo0000robe
[4.152] (Harris 1903a; Harris 1903b)
[4.153] two books, many diagrams like reed/kellogg
[4.154] based on Goold Brown, but with diagrams (Hall 1904: 36-39, 51, 57-59, 85-90, 93, 100)
[4.155] (Eastman 1904: 194-210)
[4.156] hand-drawn diagrams (White 1904: 26, 55, 60, etc.)
[4.157] redd/kellog style, but with boxes (Baker 1907: 248-254)
[4.158] Book II. An elementary English grammar, by G.L. Kittredge and S.L. Arnold (Arnold, Sarah Louise, 1859-1943; Kittredge, George Lyman, 1860-1941)
[4.159] Original does not have diagramming: (Kittredge & Arnold 1900)
[4.160] Revised edtion (1908) added appendic with diagrams (Kittredge & Arnold 1908: 338-349) with special symbols, moving upwards instead of downwards
[4.161] not yet found (Brittain 1973: 200)
[4.162] many examples of redd/kellog style diagrams (Edgar 1915: 9-66)
[4.163] graphic analysis
[4.164] https://archive.org/details/grammarbookone00perr https://archive.org/details/grammarbooktwo00perr https://archive.org/details/grammarbookthree00perr https://archive.org/details/grammarbookfour00perr
[4.165] reed/kellog with arrows! (Cross 1922: 105-116)
[4.166] https://lccn.loc.gov/28010886
[4.167] reed/kellog again (House & Harman 1931: 145-325)
[4.168] not yet found (Brittain 1973: 204)
[4.169] not yet found (Brittain 1973: 206)
[4.170] not yet found (Brittain 1973: 209)
[4.171] Original started in 1946 difficult to find. Still in use today. contains redd/kellog throughout https://en.wikipedia.org/wiki/Warriner%27s_English_Grammar_and_Composition
[4.172] “complete course” Edition from 1951 (Warriner 1951). Edition from 1969 (Warriner 1969), revised, but basically the same. Revised Edition from 1986: https://archive.org/details/englishgrammarco00holt completely revised, and number of diagramms are less, but basically still the same redd/kellog structure
[4.173] Raymond Woodbury Pence
[4.174] many reprints, later with Donald William Emery
[4.175] diagramming identical to reed/kellog (Pence 1947: 311-368)
[4.176] Describing schools in germany
[4.177] Kern is cited (p558) graphical display is presented as “normal” in Germany, example (Dale 1897: 557)
[4.178] “schema” = diagramming? (Adamson 1907: 176-177)
[4.179] cites fitch (p. 175), but does not cite inspiration for “schema”
[4.180] diagramming? (Palser & Lewis 1923) Also used in “Memorandum on the teaching of English” (1927: page 8)
[4.181] “The graphic method of analysis we have used for many years, with modifications from time to time as experience required.” (p6)
[4.182] citing: Adamson and Dale (citing Kern!)
[4.183] own graphics
[4.184] Fry (Fry 1925: 12) commenting on Palser/Lewis: “all the chief features of my method have been used by myself and a few of my pupils for something like twenty years.”
[4.185] Hypothesis that Wundt influenced Bloomfield: (Percival 1976: 234-235 gives various detailed arguments)
[4.186] not much effect in German linguistics, but see Delbrück 136ff https://books.google.de/books?id=BAQLAAAAMAAJ (no graphics though) https://doi.org/10.1515/9783111457642
[4.187] Debate about logical analysis: Wundt proposes recursive subject-predicate pairing as a model for though and language: note that this is the logical tradition of the terminology, not the grammatical!
[4.188] Wundt’s discussion of the ‘old grammarians’ (Wundt 1900: 222-224), arguing against the ‘copula-theory’
[4.189] See (Picardi 2022: 24-26) for the relation with Frege
[4.190] See (Dittrich 1902) for a direct following of Wundt in Germany, otherwise nothing found
[4.191] Alfred Dwight Sheffield (Sheffield 1912: 50-51) Grammar and Thinking (citing Wundt prominently) Brittain (1973:204) says “linguists seem to have been unaware of his work”. This is not true!
[4.192] also Alfred Dwight Sheffield (Sheffield 1929: 53)
[4.193] morphology-formula (Sapir 1921)
[4.194] (Bloomfield 1933: 161)
[4.195] (Bloomfield 1914: 60-61), cited in Percival (1976:235) clearly thinks about binary branching. Also note the Wundt-inspired psychological reasoning.
[4.196] (Ajdukiewicz 1935)
[4.197] (Pike 1943: 70), citing bloomfield
[4.198] (Wells 1947: 84)
[4.199] Francis Mikuš from Ljubljana, cf (Graffi 2001: 201)
[4.200] trees following Wundt (Mikuš 1947: 37), includes “chines boxes” diagram, without attribution.
[4.201] boxes: (Mikuš 1952: 452, 457, 460), numbers (Mikuš 1952: 453, 458), spiral (Mikuš 1952: 461), boxes with connections (Mikuš 1952: 462)
[4.202] reanalysis of Sapirs symbolic representation, boxes on page 20 (Mikuš 1953: 20)
[4.203] nothing in 1946 version! (Nida 1946)
[4.204] sentence: (Nida 1949: 87) hierarchical morphological structure p.101, both 105
[4.205] book is later called Structural Linguistics (Harris 1963[1951])
[4.206] examples of IC: my most recent plays closed down (p278-279) analysed as follows, with number of dots representing separation. Reversed bloomfield approach: more dots are “larger” separation
[4.207] T :: D . A : N² :. -s :.: V¹ . Pb : -ed
[4.208] Complete rewrite rules (in Chomskian format, harris writes them from right to left). Note the strange early past suffix insertion. This system was already introduced in (Harris 1946), but wihtout the notation for a complete sentence
[4.209] Utterance → N⁴V⁴ N⁴ → TN³ N³ → N²-s N² → AN² A → DA V⁴ → V²-ed V² → V¹Pb
[4.210] Insertion
[4.211] T = my D = most A = recent N² = play V¹ = close Pb = down
[4.212] (Fries 1952: 272) (Pike is a student of Fries, Fries might have been inspired by Pike 1943 here)
[4.213] (Bar-Hillel 1953) english explanation of (Ajdukiewicz 1935)
[4.214] (Gleason 1955: 129-131)
[4.215] second edition from 1961 extended, Chomsky mentioned in introduction (Gleason 1961). Includes generation and transformation. But still the old IC-stuff as well (Gleason 1961)
[4.216] (Gleason 1965) very open description: Ch7 on Reed/Kellog and other approaches, but also many chapters on TGG
[4.217] (Chomsky 1955)
[4.218] Chomsky (1955: 277) Note the reversal of the usage of dashes, here more alike to brackets. The dashes seem to represent some kind of intonation structure, though.
[4.219] Chomsky (1955: 316b) Similar to Pike/Fries/Hockett
[4.220] Chomsky cites Bar-Hillel (1953: 57)
[4.221] also in (Chomsky 1975: 229, 258)
[4.222] (Hockett 1958: 151-155, 178-180, 188-189)
[4.223] (Francis 1958) chinese boxes. Stewart (Stewart 1976) reports from personal communication with Francis, that the chinese-box diagrams were inspired on the graphical display in (Fries 1952)
[4.224] (Nida 1960) added images to 1943 dissertation, which did not have images
[4.225] Nida (1973[1960]) writes his thesis in 1943, which is published in 1960. There do not seem to be trees in his thesis, but he Immediate Constituents might imply trees (but difficult to follow). In the 1960 edition he adds a long list of trees. But this is of course after Chomsky 1957.
[4.226] In the preface of the 1943: 30 dissertation he cited Sheffield 1912 (though not with respect to structural analysis)
[4.227] Note the combination of dependency and constituency!
[4.228] (Gammon 1963), completely ignoring chomsky?
[4.229] https://books.google.de/books?id=YskUAAAAIAAJ https://archive.org/details/towardsscienceof0000euge
[4.230] summary of different equivalent graphics (Nida 1964: 58-62)
[4.231] chinese boxes (Stageberg 1965: 263) Fries-scheme of IC (Stageberg 1965: 263-273)
[4.232] added chapter on TGG from different author. seems to be added late in the process of making the book.
[4.233] https://lccn.loc.gov/10028710 https://lccn.loc.gov/10028727 https://lccn.loc.gov/10033615
[4.234] (Brittain 1973: 31)
[4.235] completely idiosyncratic system of letters and numbers. No references whatsoever (Roberts 1956).
[4.236] (Smith 1957) idiosyncratic system with symbols
[5.1] Carl Friedrich Nägelsbach (1806-1859) was professor for classical studies in Erlangen. He wrote various commentaries on classical texts and a didactic explanation of Latin style, called Lateinische Stilistik für Deutsche, ein sprachvergleichender Versuch (Nägelsbach 1846). The book saw various editions up to a ninth edition in 1905, prepared by Iwan Müller. This ninth edition has been reprinted up to at least 1980.
[5.2] In the Lateinische Stilistik Nägelsbach uses Periodenbilder to analyse latin constructions, citing both Lehmann (Nägelsbach 1846: 281) and Götzinger (Nägelsbach 1846: 282) as inspiration. He simplifies the structure of the Periodenbild to be able to write them on a single line, using slashes instead of horizontal braces, as illustrated in Figure 5.1 (Nägelsbach 1846: 286).
[5.3] Another innovation proposed by Nägelsbach is to use brackets for relative modification at the front of the sentence. Specifically, he notes that in Latin it is quite common to have multiple Vordersätze, i.e. clauses in front of the main clause, as indicated by the colon in Lehmann’s notation. Because of the ordering of modification, Nägelsbach proposed Periodenbilder like “a:(b:A)” and even “a:(b:[c:A])” for reverse-nested frontal clauses (Nägelsbach 1846: 286, 289).
[5.4] Hermann August Menge (1841-1939) was a teacher and is widely known as the translator of the infamous German Menge-Bibel. Based on his practical experience as a teacher he wrote textbooks for Latin and Greek. The Latin textbook was still in use far into the 20th Century, though in a slightly revised version (e.g. a 1955 edition is cited in Langholf 1969: 10). I will use here the sixth edition, which is the earliest edition I have been able to access (Menge 1890). The prefaces to the earlier editions do not mention any changes as far as the graphical analysis is concerned, so I assume that they were already present in the first edition of 1873.
[5.5] Menge (1890: 357-364) uses a graphical analysis of the Periode in the tradition of Lehmann (see Section 3.2.8). However, he uses a forward slash instead of a horizontal brace, just like Nägelsbach (see Section 5.1.1). In the preface to the first edition Menge cites the influence of many different grammars of Latin (Menge 1890: v), but not Nägelsbach. However, Nägelsbach is mentioned in passing at various times throughout the book (e.g. Menge 1890: v, 342). So it seems very likely that the use of the forward slashes is inspired by Nägelsbach.
[5.6] Ernst Hendrik Ferdinand Schulze (1842-1911) was a teacher and school director, first in St. Petersburg and later in Bad Homburg vor der Höhe. Biographical details are available in Schulze (1901: 46-47). A good overview is also available on https://de.wikipedia.org/wiki/Ernst_Schulze_(Historiker), accessed 7 January 2025. He wrote an advanced Latin textbook Adiumenta Latinitatis. Grundzüge des lateinischen Stils in Verbindung mit Übersetzungsstücken für die oberste Stufe des Gymnasiums (Schulze 1883).
[5.7] In this book he used very simple Periodenbilder for the analysis of Latin sentences (Schulze 1883: 116-122). He only uses the letters A for main clause, a for subordinate clause, and α for secondary subordinate clause. Specifcally, he used this graphical analysis to explain differences between German and Latin sentence structure. The regular Latin word order would be a α A, as shown in (5.1). However, the Latin word order is quite flexible, and many different orders are possible. Schulze then gives the table shown in Figure 5.3 with the orders that are possible in Latin, but not in German.
(5.1) | a. | a facias me certiorem α quid istic agatur A velim. |
b. | faci-as | me | certior-em | quid | istic | ag-atur | velim. | ||
make-2sg | 1sg.obj | surer-acc.sg | what | there | do-pass.3sg | wish.1sg | |||
‘I would like you to inform me what is going on there.’ |
[5.8] Heinrich Gloël (1855-1940) was a teacher and literary scholar, first in Wesel and and later in Wetzlar. He started as a classical scholar, writing a disseration about the Greek play Hippolytos by Eurypides and teaching classical languages. Later in life he specialized in critical evaluations of texts of Goeths (more biographical information in Flender 1983).
[5.9] In an article about the role of translation from Greek to German in secondary education, Gloël discussed various examples from classical Greek and the problems with their translation into German (Gloël 1895). The final example he discusses is an excerpt from Plato’s Crito (Gloël 1895: 67-76). For two complex sentences with many subordinated clauses in this excerpt, Gloël uses Satzbilder to illustrate their syntactic structure. For example, the Satzbild in Figure 5.4 is an analysis of the sentence in (5.2 a). I have added the letters from Gloël’s Satzbild into this example, something which Gloël assumes that the reader is able to do. The translation in (5.2 b) is by Harald Fowler. Plato in Twelve Volumes, Vol. 1 translated by Harold North Fowler; Introduction by W.R.M. Lamb. Cambridge, MA, Harvard University Press; London, William Heinemann Ltd. 1966. Available online at https://www.perseus.tufts.edu/.
(5.2) | a. | (A) ἢ πρὸς μὲν ἄρα σοι τὸν πατέρα οὐκ ἐξ ἴσου ἦν τὸ δίκαιον καὶ πρὸς δεσπότην, (a) εἴ σοι ὢν ἐτύγχανεν, (b) ὥστε (β) ἅπερ πάσχοις ταῦτα (b) καὶ ἀντιποιεῖν, (1) οὔτε κακῶς ἀκούοντα ἀντιλέγειν (2) οὔτε τυπτόμενον ἀντιτύπτειν (3) οὔτε ἄλλα τοιαῦτα πολλά· (B) πρὸς δὲ τὴν πατρίδα ἄρα καὶ τοὺς νόμους ἐξέσται σοι, (c) ὥστε, (γ) ἐάν σε ἐπιχειρῶμεν ἡμεῖς ἀπολλύναι (p) δίκαιον ἡγούμενοι εἶναι, (c) καὶ σὺ δὲ ἡμᾶς τοὺς νόμους καὶ τὴν πατρίδα (δ) καθʼ ὅσον δύνασαι (c) ἐπιχειρήσεις ἀνταπολλύναι, (1) καὶ φήσεις ταῦτα ποιῶν (2) δίκαια πράττειν, (p) ὁ τῇ ἀληθείᾳ τῆς ἀρετῆς ἐπιμελόμενος; (Gloël 1895: 67, citing Plato Crito. Clause-symbols are not given in the original, but inserted here to coincide with the Satzbild below) |
b. | (A) There was no such equality of right between you and your father or your master, (a) if you had one, (b) so that (β) whatever treatment you received (b) you might return it, (1) answering them if you were reviled, (2) or striking back if you were struck, (3) and the like; (B) and do you think that it will be proper for you to act so toward your country and the laws, (c) so that (γ) if we undertake to destroy you, (p) thinking it is right, (c) you, to our laws and our country, (δ) so far as you are able, (c) will undertake in return to destroy them (1) and will say that in doing this (2) you are doing right, (p) you who really care for virtue? Translation by Harold North Fowler with a few very minor edits, available online at https://scaife.perseus.org/reader/urn:cts:greekLit:tlg0059.tlg003.perseus-grc2:51/?right=perseus-eng2. |
[5.10] https://de.wikipedia.org/wiki/Josef_Golling
[5.11] 2nd edition (Golling 1900), but first edition from 1892 is nowhere to be found 5th edition 1922
[5.12] cites Nägelsbach and follows his abbreviations, e.g. vi-viii
[5.13] Severin Carl Olfert Fischer Broberg (1822-1900) was a danish author and occasional lecturer at the Universities of Randers and Odense. He wrote a Danish language guide for French speakers, called Manuel de la langue danoise (Broberg 1882). To explain the Danish sentence structure he used a special graphical technique of an indented table (Broberg 1882: 95-102). There is no indication towards any inspiration that might have led Broberg to propose this kind of analysis.
[5.14] Repeated here is one of Broberg’s Danish example sentence in (5.3 a), with his own French translation (5.3 b). An glossed version with English translation is added in (5.3 c). Broberg’s graphical analysis is shown in Figure 5.5. The basic sentence Datter viste et Maleri ‘Daughter showed a painting’ is shown in bold typeface at the left side of the tabular display. All modifier phrases are shown below their head phrases, indented to the right. The order of the phrases from top to bottom reflects their modification order. For example, Datter ‘daughter’ is first modified by ældste ‘oldest’ and then by Generalens ‘of the general’. However, the order of these elements in a Danish sentence is reversed: the real order in which the words occur in the Danish sentence is indicated by the numbers in the table.
(5.3) | a. | Generalens ældste Datter viste, for at give Samtalen en anden Vending, en af Gjæsterne et fra Kjøbenhavn ankommet, en læsende Hyrde forestillende Maleri. | |
b. | La fille aînée du général, pour faire diversion à l’entretien, montra à l’un des convives un tableau venu de Copenhague et représentant un berger lisant. |
c. | Generalens | ældste | Datter | viste, | for | at | give | Samtalen | en | anden | Vending, | en | af | Gjæsterne | et | fra | Kjøbenhavn | ankommet, | en | læsende | Hyrde | forestillende | Maleri. | ||
general | oldest | daughter | showed | for | to | give | talk | an | other | turn | one | of | guests | the | from | Copenhague | arrived | a | reading | shepherd | showing | painting | |||
‘The general’s eldest daughter, to give the conversation a different turn, showed one of the guests a painting that had arrived from Copenhagen, depicting a reading shepherd.’ |
[5.15] Otto Jespersen (1860-1943) was professor of English at Copenhagen. Jespersen worked on a wide variety of linguistic topics and throughout his long life he prepared a massive 7-volume Modern English grammar on historical principles. For most of his career he did not apply graphical or algebraic methods to linguistic analysis. There are just a few incidental symbolic representation in Sprogets logik ‘The logic of language’ (Jespersen 1913: 80-81) and a single graphical representation in De to hovedarter av grammattiske forbindelser ‘The two main types of grammatical connections’ (Jespersen 1921: 18). Only towards the end of his life he published Analytic syntax (Jespersen 1937), in which he laid out a completely symbolic framework for linguistic analysis. It feels like an idea he probably had been working on for a long time, but only published late in his career.
[5.16] Jespersen proposed various structural principles for linguistic analysis, above all his notion of rang ‘rank’ (first introduced in Jespersen 1913: 31). The basic idea of rang is that there are different levels of modification in language. Jespersen uses roman capital letters I, II, III for these levels. He named the first three levels as overled, adled and underled (Jespersen 1913: 31) or in a different publication primær, sekundær and tertiær (Jespersen 1921: 3) and in English principal, adjunct and subjunct (Jespersen 1913: 31). He clearly intends there to be more levels, but he claims without much argumentation that further levels are inherently the same as the third level (Jespersen 1921: 3).
[5.17] Jespersen considers the first two levels to be the most interesting: Da de tertiære elementer ikke gir anledning til videre bemærkninger, skal jeg nu helt forlade dem for at holde mig til forbindelserne mellem primære og sekundære elementer ‘Since the tertiary elements do not warrant further comment, I must now leave them entirely in order to stick to the connections between primary and secondary elements’ (Jespersen 1921: 4). Jespersen goes into great detail about the many different ways level I and II occur in language. Curiously, he quickly stops using the letter I and II and changes to P and S instead. These letters are reminiscent of the terms “predicate” and “subject”, though Jespersen does not want to use these terms:
den består av to led, som vi for at undgå forvexlinger ikke tør kalde subjekt og prædikat, men foreløbig betegner ved bogstaverne S og P. (Jespersen 1921: 7)
(‘it consists of two parts, which, to avoid confusion, we dare not call subject and predicate, but for the time being denoted by the letters S and P.’)
[5.18] The abstract use of the letters S and P is probably an influence from Wundt (see Section 4.7.1) and Sheffield (see Section 4.7.2). Jespersen cites both these authors at various places throughout his work, although not directly in connection to the abbreviations S and P. The impression arises that Jespersen had better used different terms like “dependent” and “head”, but was stuck in the Wundtian tradition.
[5.19] The only graphical analysis that I have been able to find in Jespersen’s work also points to an influence from Wundt. Jespersen uses the letters S and P recursively, similar to Wundt, in a footnote citing an incomplete quotation from Quintus Curtius Rufus’ Historiae Alexandri Magni (5.4). He then explains that vi det ejendommelige at S i den ene nexus selv er en nexus ‘we have the peculiarity that S in one nexus is itself a nexus’ (Jespersen 1921: 18). The subsequent graphical represenation of this analysis is shown in Figure 5.6.
(5.4) | Latin (Jespersen 1921: 18, quoting Quintus Curtius Rufus) | ||||||||
Alexander, | audito | Dareum | movisse | ab | Ecbatanis, | fugientem | insequi | pergit | |
Alexander | hear.part | Darius.akk | move.inf | from | Ecbatana | fleeing man | follow.inf | continue.3sg | |
‘Alexander, hearing that Darius had moved away from Ecbatana, continued to pursue the fleeing man.’ |
[5.20] At the end of his long life Jespersen published the book Analytic syntax (Jespersen 1937), in which he laid out a completely symbolic framework for linguistic analysis, leaving the simplistic S/P approach behind.. The approach in Analytic syntax is algebraic, using symbols and brackets to analyse linguistic utterances. He does not draw any graphical representations, although Francis (1989: 83) shows that is clearly possible to interpret the bracketing as a constituent tree. Also, Jespersen (1937: 83) includes a graphical syntactic analysis in his book, citing Fry (see Section 4.6.4), so he was clearly aware of the possibility to use images to analyse language. However, for some reason Jespersen preferred an algebraical approach, citing Stöhr (see Section 3.4.3), Sapir (see Section 4.7.3), Bologne (see Section 5.7.3) and Brøndal (see Section 5.2.5) as having similar approaches.
[5.21] Jespersen uses letters, superscripts and brackets to represent syntactic structure, and, being the detailed grammarian that he has been for his whole life, he presents many pages with hundreds of examples of this system for a plethora of different grammatical phenomena from a dozen different languages (Jespersen 1937: 6-84). This is hands-down the most well-documented system of all approaches discussed in this book. A more in-depth discussion of Jespersen’s approach can be found in Francis (1989) and Cigana (2020). As an illustration of his approach an example is shown here in Figure 5.7 for an example sentence taken from Fry (1925: 64).
[5.22] Louis Trolle Hjelmslev (1899-1965) was a linguist at the university of Copenhagen and co-founder of the Lingvistkredsen ‘Linguistic Circle of Copenhagen’ with Brøndal (see Section 5.2.5). Hjelmslev developed an intricate system of linguistic analysis under the label glossomatik, but he does not appear to have applied this to the analysis of longer stretches of text. There are many graphics in his work, but they mostly deal with paradigmatic structures. A rare syntactic visualisation is the rather trivial example shown in Figure 5.8 (Hjelmslev 1928: 138) for the Latin sentence in (5.5 a). A few years later he adds another trivial visualisation, as shown in Figure 5.9 (Hjelmslev 1935: 52) for the Latin example in (5.5 b). These visualisations are probably better analysed as showing government/agreement relations, and not as a real syntactic sentence analysis.
(5.5) | a. | Latin (Hjelmslev 1928: 138) | |||
dā‑b‑ō | frātrī | librum | |||
give‑fut‑1sg | brother.dat | book.akk | |||
‘I will give (my) brother a book.’ |
b. | Latin (Hjelmslev 1935: 52) | ||||
rosa | est | pulchra | |||
rose.nom | be.3sg | beautiful.nom | |||
‘The rose is beautiful.’ |
[5.23] In Liber americorum for Jespersen’s 70th birthday, closely following Jespersen (see Section 5.2.2)
Sage ich: “Der Knabe schlägt den Hund”, so haben wir hier zwei Gedanken 1) der Knabe schlägt, dadurch 2) der Hund wird geschlagen. (Hammerich 1930: 312)
(‘When I say “the boy hits the dog”, then we have two thoughts 1) “the boy hits”, and therefore 2) “the dog is begin hit”.’)
[5.24] Rasmus Viggo Brøndal (1887-1942) was a professor at Copenhagen and a co-founder of the Lingvistkredsen ‘Linguistic Circle of Copenhagen’ (biographical information is available in Hallon 1989: xvi-xxvi). In his books Ordklasserne: partes orationis. Studier over de sproglige kategorier (Brøndal 1928) and more specifically in Morfologi og syntax (Brøndal 1932; translated in Hallon 1989) Brøndal develops a highly idiosyncratic system for linguistic analysis based on symbolic combinations of the letters D, d, R, r, as summarised in the fold-out appendix to the Ordklasserne (Brøndal 1928), shown here in Figure 5.11.
[5.25] In his books, Brøndal applies combinations of these letters (extended with colons and italic typesetting) for all kinds of different linguistic phenomena. Unfortunately, his examples for the application of this system to actual linguistic data are very short and difficult to interpret. For example, there are no real analyses of longer sentences in the discussion about syntax (Brøndal 1932: Chapter 2). Togeby (1989) proposes an interpretation of Brøndal ideas in more modern terms and visualisations. Specifically, he argues that Brøndal actually describes a hierarchichal constituency structure for sentences as shown in Figure 5.12 for the example in (5.6 a). For the internal structure of noun phrases, Brøndal (1932: §70) uses roman numerals I through VI, which Togeby (1989: 110) interprets as a templatic structure, shown in Figure 5.13 for the examples in (5.6 b). Such a templatic structure only becomes mainstream with the work of Diderichsen (see Section 5.2.7).
(5.6) | a. | … | at | dreng-en | ikke | omhyggeligt | har | malet | porten | rød. | |
… | that | boy-def | not | carefully | has | painted | gate | red. | |||
‘… that the boy has not carefully painted the gate red.’ |
b. | de | mange | uartige | små | børn | du | så | i går | ||
the | many | naughty | little | child.pl | you | see.past | yesterday | |||
‘the many naughty little children that you saw yesterday’ |
[5.26] Aage Hansen (1894-1983) studied with Otto Jespersen in Copenhagen and worked at that university for the rest of his career. He was a central author in the preparation of two major encyclopaedic works about the Danish language, namely the dictionary Ordbog over det danske Sprog and the grammar Moderne Dansk (more biographical information in Hjorth & Danske 2025). Earlier in his career Hansen wrote a rather programmatic book about language analysis, called Sætningen og dens led in moderne dansk ‘sentences and its parts in modern danish’ (Hansen 1933). This book is dedicated to Jespersen and contains various graphical analyses. However, the graphics do not play a central role in his argumentation and are only used sporadically as illustrations.
[5.27] Hansen uses examples from danish literature throughout his book, and he is particularly precise in referencing these examples. The examples in (5.7), used for the graphics shown below, are from the author Karl Larsen. In contrast, Hansen does not cite much scientific literature, so it remains unclear how his diagrams originated. However, he regularly mentions his teacher Jesperson, who cites (e.g. in Jespersen 1913) the work of Sheffield (see Section 4.7.2) and Wundt (see Section 4.7.1). So the diagrams of Hansen are probably related to the immediate constituent tradition developing in the USA at the time Hansen is writing this book.
[5.28] The whole book is centered around the assumption that a sentence consists of two parts, which Hansen called the A-led ‘part A’ and B-led ‘part B’ (Hansen 1933: 23). He explicitly explains that this division is not equivalent to the subject/predicate devision, but more alike to a topic/comment opposition or the classical onoma/rhema distinction (Hansen 1933: 69).
en meddelelse består af to nødvendige elementer: en angivelse af den størrelse, man vil meddele noget om, og en angivelse af selve det, man vil meddele om denne størrelse. (Hansen 1933: 22)
(‘A message consists of two necessary elements: a statement of the quantity you want to communicate, and a statement of what you want to communicate about this quantity.’)
[5.29] The graphical illustrations in his book only deal with the internal structure of the part B, although this might be a coincidence due to the sparse usage of the illustrations. Hansen uses three different graphical approaches, all intended to illustrate hierarchical modification of constituents. The first approach is a “chinese boxes” or “Matryoshka dolls” visualisation in which modifiers are hierachically organised by drawing enclosing circles or boxes. One of his drawings is shown here in Figure 5.14 for a small part of the example (5.7 a). Hansen’s second approach is particularly innovative, and I have not seen anything like this before or since. As shown in Figure 5.15 Hansen draws a Planetskemaet ‘planetary scheme’, in which a modifier is metaphorically compared to a planet or moon circling around another celectial body.
[5.30] Finally, he also uses a more ‘regular’ visualisation using hierarchically ordered brackets, with a Roman numeral I for the modified element and a Roman numeral II for the modifier. This approach is shown in Figure 5.16 for part of the sentence (5.7 b). The use of Roman numerals is probably taken from Jespersen’s rang notation (Jespersen 1913: 31). The hierarchical brackets are also found in Jespersen’s work (see Section 5.2.2), but Jespersen never seems to have applied this visualisation as consistently as Hansen does here (see Section 5.2.2).
(5.7) | a. | Danish (Hansen 1933: 92, citing Karl Larsen) | |||||||||||
saa | har | vi | her | en | lille | Samling | ganske | særligt | fremragende | Videnskabsmænd. | |||
so | here | we | have | a | small | collection | very | special | outstanding | scientists | |||
‘So here we have a small collection of very outstanding scientists.’ |
b. | Danish (Hansen 1933: 105, citing Karl Larsen) | ||||||||||||||||
Han | udmærk‑ede | sig | ganske | særligt | ved | en | naiv | Mangel | paa | Kendskab | til | alle | praktiske | Ting. | |||
he | mark-past | refl | very | special | by | a | naive | lack | of | knowledge | for | all | practical | matters | |||
‘He was particularly distinguished by a naive lack of knowledge of all practical matters.’ |
[5.31] https://de.wikipedia.org/wiki/Paul_Diderichsen
[5.32] second edition seems identical to first edition from 1947
[5.33] Feldertheorie (Diderichsen 1957), giving nested schema like hansen (Diderichsen 1957: 146) does not cite drach (1935) uses term “chinese boxes”
[5.34] kinesisk Æskesystem ‘Chinese box-system’ (146)
[5.35] “Disse Led markerer da Afsnit inden for den større Helhed, som vi vil kalde „Felter”; inden for Felterne kan der saa atter markeres en Række „Pladser”.” (151)
[5.36] These parts mark sections within the larger whole, which we will call “fields”; within the fields a series of “places” can then again be marked.
[5.37] 1964 English translation???
[5.38] details about history: https://da.wikipedia.org/wiki/Sætningsskema
[5.39] example p. 162, 167 (only later editions), 186
[5.40] He knew about Hansen, but never cited it. acknowledgement in 1964 Festschrift for Hansen (Diderichsen 1966: 364).
[5.41] Anshelm Fredrik Wilhelm Lindwall (1859-1931) was a teacher in Stockholm. Some biographical information is available on the swedish Wikipedia https://sv.wikipedia.org/wiki/Wilhelm_Lindwall, accessed 11 January 2025. In 1903 he writes a short paper with the title Några ord om satsanalys ‘a few words about sentence analysis’ (Lindwall 1903). This paper starts out as a review of a grammatical textbook by Rebbe & Fischer, but in the second part of this review Lindwall proposes a graphical display for the exaplanation of syntax med tillhjälp af svarta taflan ‘with the help of the blackboard’ (Lindwall 1903: 137). There is no indication given by Lindwall about the inspiration for his drawings.
[5.42] Lindwall uses two different visualisation, one for the internal structure of single-clause sentences and another one for the structure of multi-clause sentences. His single-clause analysis is illustrated here with Figure 5.17 for the example in (5.8 a). The predicate is depicted as a double horizontal line, while the subject is a vertical line upwards, the object a vertical line downwards and prepositional phrases are shown as the diagonal line downwards. Adverbials are crosses on the line, adjectives are diagonal slashes through the line (not shown in this example).
[5.43] For multi-clause sentences, each clause is depicted by a line, as shown here in Figure 5.18 for the example in (5.8 b). In these graphics, all lines are drawn in the same order as they occur in the sentence. Main clauses are horizontal lines, while subordinate clauses are vertical lines. Secondary embedded subordination is indicated by diagonal lines. So, the example in Figure 5.18 has a main clause (horizontal: jag hoppas), a subordinate clause (vertical: att du kommer), then a secondary subordinate clause (slanted down rightwards: när du slutat arbetet) and finally even a tertiary subordinate clause (slanted down leftwards: hvarmed du varit sysselsatt). Lindwall proposes a few more techniques (not shown here), e.g. a plus-symbol for coordination and dashed lines for non-finite clauses.
(5.8) | a. | Man-nen | gaf | genast | en | slant | ât | den | fattige. | |
man-def | gave | immediately | a | coin | to | the | poor | |||
‘The man immediately gave a coin to the beggar.’ |
b. | jag | hoppas, | att | du | kommer, | när | du | slutat | arbetet, | hvarmed | du | varit | sysselsatt. | ||
I | hope | that | you | come | when | you | stop | work | with.which | you | are | busy | |||
‘I hope you will come when you have finished the work with which you were busy.’ |
[5.44] Karl Frederik Natanael Beckman (1868-1946) was a high school teacher and later professor of Scandinavian languages at Gothenburg University (more biographical details are available in Friesen 1922). He wrote many different analyses of Swedish, both synchronically and diachronically. His Svensk språklära för den högre elementarundervisningen (Beckman 1904) has been widely used and has remained in print at least until a 9th edition in the 1960s. He also prepared two reduced versions of this work, namely the Svensk språklära för elementarundervisningen (Beckman 1908) and the Svensk språklära för folkskolan (Beckman 1912). These do not seem to add anything new and will not be further discussed here.
[5.45] Beckman (1904: 273) cites Kern (see Section 3.3.24) and Lindwall (see Section 5.3.1) as inspiration for his graphical display, which he calls a karta ‘map’. He uses a balanced display with subject and object at the top linked by a line (Beckman 1904: 194-198, 202, 219), explicitly distancing his approach from Kern’s verb-centric display (Beckman 1904: 273). An example for a single-clause sentence is shown in Figure 5.19 for the sentence in (5.9 a). This figure is actually a compositive from parts that are explained separately by Beckman. However, Beckman invites the reader to produce this combined picture, to which I have happily obliged: Du kan nu själf rita en karta öfver hela satsen ‘You can now draw a map of the entire sentence yourself’ (Beckman 1904: 195).
[5.46] Dependencies are linked by upwards-pointing arrows. Multiple dependencies to the same head are indicated by horizontal brackets. Somewhat confusingly, sometimes the arrows are missing, as with fädernesland ‘homeland’ and från välde ‘from the reign’ in Figure 5.19. It is unclear whether this is an error, or whether Beckman intended to omit the arrows for arguments. For multi-clause sentences Beckman simply draws separate figures for each clause, and indicates with the word bisatz ‘subordinate clause’ where the subordinate clause has to be inserted into the matrix clause. An example of a multi-clause sentence in shown in Figure 5.20 for the example in (5.9 b).
(5.9) | a. | Vår | store | och | ädle | konung | Gustaf | Vasa | har | befriat | vårt | gamla | fädernesland | från | danskarnas | tryckande | välde. | |
our | great | and | noble | king | Gustaf | Vasa | has | freed | our | old | homeland | from | Danish | oppressive | reign | |||
‘Our great and noble King Gustaf Vasa has liberated our old homeland from the oppressive rule of the Danes.’ |
b. | Den | af | er, | som | kan | svara | på | denna | fråga, | skall | få | en | slant. | ||
dem | of | 2pl | rel | can | answer | on | dem | question | shall | get | a | coin | |||
‘Whoever of you can answer this question will receive a coin. (Beckman 1904: 194, 202)’ |
[5.47] Adolf Gotthard Noreen (1854-1925) was a professor for Germanic languages at Uppsala University (More biographical information in Elmevik 1990). His work was of great influence on all of scandinavian linguistics, but he does not seem to have used graphical analyses for sentence structure. However, he used an interesting graphical approach to morphological structure in the 7th volume of his magnum opus Vårt språk (Noreen 1906).
[5.48] Noreen uses dashes to separate morphemes, but additionally the number of dashes is crucial for the hierarchical morphological structure. The more dashes, the more distant the morphological connection: jag låter den primära sammansättningens leder skiljas af ett bindestreck, den sekundäras af två osv. ‘I separate the primary compounds by a hyphen, the secondary by two, etc.’ (Noreen 1906: 32-33). Two of his examples are shown in (5.10) and a modern rendition of the intended hierarchical structure is displayed in Figure 5.21.
(5.10) | a. | sprit-handels---aktie--bo-lags----ord-förande | |
liquor-trading---share--asset-alliance----word-carrier | |||
‘chairman of a liquor trading company’ |
b. | kropps-arbetare--för-säkrings---kommitté----ut-låtandet | ||
body-worker--for-safety---committee----out-letting | |||
‘statement of the manual worker’s safety committee’ |
[5.49] Carl Rebbe (1862-1948) was a teacher in Gothenburg (some basic biographical information is available in Sjögren 1949). In 1912 he wrote the Svensk språklära för realskolan ‘Swedish language-teaching for middle school’. This book has been extremely popular and remained in print in various editions and revisions until the 1970s, revised by either Nils Ivan, Helge Gullberg and/or Birger Bjerre. I have only been able to access the tenth edition from 1935 (Rebbe 1935). Starting in 1917 Rebbe also prepared a frequently reprinted kortfattad ‘concise’ version of the Svensk språklära. And in 1902 he published a collection of examples called Exempelsamling till inöfvande af allmänna satsläran och skiljetecknens bruk ‘Collection of examples for practicing general syntax and the use of punctuation marks’. Walter Fischer revised this collection in 1914 with a minor change in title (svenska satsläran instead of allmänna satsläran). All these different textbooks and their numerous editions need a more in-depth investigation and comparison.
[5.50] In the 1935 version, there is a single page with various examples of the graphical analysis of single-clause sentences (Rebbe 1935: 188). An example is shown in Figure 5.22 for the example in (5.11). The pointing arrows are clearly similar to the approach from Beckman (see Section 5.3.2). The other examples also show various details that indicate that this visual display was inspired by Beckman (e.g. adverbials pointing to the line connecting subject and predicate).
(5.11) | Swedish (Rebbe 1935: 188) | |||||||
Eriks | far | har | givit | honom | en | dyrbar | färglåda. | |
Erik‑gen | father | has | given | 3sg.obj | art | precious | paintbox | |
‘Erik’s father has given him a precious paint box.’ |
[5.51] Eemil Nestor Setälä (1864-1935) was a well-known Finnish politician and author of the Finnish declaration of independence. I thank Fred Karlsson and Matti Miestamo for their assistance with the Finnisch language and linguistic tradition. However, he was also a professor for Finnish language and literature at Helsinki University and among many other grammatical works he wrote a little booklet Suomen kielen oppikirja ‘Finnisch language textbook’. The first edition appeared in 1898 and has been reprinted various time. The version from 1925, that I have consulted, does not have any graphical analyses (Setälä 1925).
[5.52] Kaarlo Nieminen (1887-1955) was a Finnish language teacher and acting professor of Finnish literature at Helsinki University for a few years. After the death of Setälä he revised the Suomen kielen oppikirja in 1939, which was reprinted many times. In this revision Nieminen added a graphical analysis for the example sentence in (5.12) as shown in Figure 5.23, cited here from a later reprint (Setälä & Nieminen 1946: 68). It is unclear whether this approach with arrows for modification has any precursors, or whether this is an original idea from Nieminen:
Lauseenosien keskinäisiä suhteita voidaan havainnollistaa piirroksen avulla. Jos merkitsemme subjektin paksulla pystysuoralla viivalla, predikaatin siitä lähtevällä paksulla vaakasuoralla viivalla, kumpaisenkin määräykset niihin vinosti suuntautuvilla ohuemmilla viivoilla sekä määräysten määräykset taas niihin suuntautuvilla viivoilla. (Setälä & Nieminen 1946: 68)
(‘The mutual relations of the parts of a sentence can be illustrated by means of a drawing. If we mark the subject with a thick vertical line, the predicate with a thick horizontal line extending from it, the modifiers of each with thinner lines directed at them diagonally, and the modifiers of the modifiers again with lines directed at them.’)
(5.12) | Mattila-n | pienet | pojat | luke-vat | hyvin | ahkerasti | vaikeita | läksyjä-nsä |
Mattila-gen | little | boys | study-3pl | very | diligently | difficult | homework-3poss | |
‘The little boys from Mattila are studying their difficult homework very diligently.’ |
[5.53] Aulis Onni Ojajärvi (1913-1968) was a dialectologist and language teacher. Some biographical information is available online at http://www.norssit.fi/sivut/5_5_norssin_ojajarvi.php, accessed 13 January 2025. In 1964, after the death of Kaarlo Nieminen, he once again revised the Suomen kielen oppikirja. The version from 1970, that I have consulted, is rather different from earlier versions and contains completely different graphical analyses (Setälä, Nieminen & Ojajärvi 1970). The structure from the examples sentence (5.13) is illustrated in two different ways in Figure 5.24. The upper version is strongly reminiscent of the visualisation of Beckman (see Section 5.3.2). The lower display is structurally identical, but the words are placed in their regular sentence order.
(5.13) | Juuri | kaupungi-sta | saapunut | setä | anto-i | minu-lle | pussi-n | makeis-i-a. |
just | town-ela | arrived | uncle | give-pst | me-all | bag-gen | sweets-pl-par | |
‘An uncle who had just arrived from town gave me a bag of sweets.’ |
[5.54] For complex sentences with subordinate clauses, Setälä, Nieminen & Ojajärvi (1970: 97-98, 104-105) use a variant of the musical-score approach. Main clauses (päälause) are placed on top and subordinate clauses (sivulause) are graphically lowered depending on the depth of embedding. The example in (5.14) is graphically analysed in Figure 5.25.
(5.14) | Jos | auta-t | minu-a, | kun | tarvit-sen | apu-a, | anna-n | sinu-lle | tämä-n. |
if | help-2sg | me-par | when | need-1sg | help-par | give-1sg | you-all | this-gen | |
‘If you help me, when I need help, I will give you this.’ |
[5.55] Antal Lonkay (1827-1888) was a teacher in Pest, Hungary (some biographical information is available in Benda 1971). He wrote a German grammar for speakers of Hungarian, called Gyakorlati Nemet Nyelvtan az ausztriai birodalombeli elemi tanodak masodik es harmadik osztalyainak szamara ‘Practical German grammar for the second and third grades of elementary school students in the Austrian Empire’ (Lonkay 1855).
[5.56] In this grammar Lonkay used a templatic table for German sentence structure (see Good 2016 for the linguistic notion of a template). Most interestingly, Lonkay distinguishes specific positions for the finite verb at the second position and for non-finite parts at the end. The table is shown here in Figure 5.26 (Lonkay 1855: 175). This syntactic analysis is very similar to what today in German grammar is called a “topological model”, which is commonly attributed to Drach (1937, see Section 3.5.1) and/or Diderichsen (1946, see Section 5.2.7). However, as shown by this example from Lonkay, this approach is much older, possibly even dating back to Becker (1830, see Section 4.3.1). There is no indication in the book for any inspiration that Lonkay might have had for this approach.
[5.57] The first position in this table is called alany ‘subject’ and the rest allitmány ‘predicate’. Lonkay also explicitly discusses examples with non-subjects in first position on the next page as deviating from the structure shown in this table. The section called ‘predicate’ is separated into the following columns:
[5.58] Lonkay explains in a bit more detail which elements are placed in this final position, which today is commonly called the Rechte Satzklammer ‘rightmost sentence bracket’. This proposal to combine non-finite verbs together with separable preverbs (‘prepositions’) into a single slot at the end of a sentence is, for example, not yet found in Drach (1937).
… végre az igenév vagy a határtalan mód, vagy bezárhatja még a mondatot az igének elülröl elszakított s hátratett elöljárója. (Lonkay 1855: 175)
(‘… finally the participle or indefinite mood, or the sentence may be closed by the preposition of the verb that has been separated from the front and placed behind.’)
[5.59] Sámuel Brassai (1800-1897) was a polymath in the Austro-Hungarian empire, who worked most of his life at the university of Klausenburg (today Cluj, Rumania). He published in Hungarian about a wide variety of subjects, including linguistics (more details about his linguistic work in Imrényi & Vladár 2020).
[5.60] Of particular interest is his Paraleipomena kai diorthoumena. A mit nem mondtak s a mit roszul mondtak a commentatorok Virg. Aeneise II. könyvére ‘Omissions and corrections: What was not said and what was wrongly said by the commentators on Book II of Virgil’s Aeneid’ (Brassai 1873). As the title clearly states, this book is a discussion of classical literature. The graphical analyses are inside a (long) footnote, explaining that adverbial modifiers can have different interpretations, depending on which word they modify. The graphical illustrations are thus only used as an aside, it is not the central topic of the discussion at all.
[5.61] To exemplify the problem of adverbial modification, Brassai chooses a sentence rather randomly from classical literature, ending up with a sentence from Ovid, shown here in (5.15 a) with a poetic translation in (5.15 b). Ovid’s sentence structure is infamous for being particularly unwieldy. Out of context, the adverbial phrase imbre per indignas usque cadente genas ‘a shower (of tears) falling on innocent cheeks’ could either modify the verb tenebat ‘she held him’ (Figure 5.27 I), the accusative flentem ‘the weeping one’ (Figure 5.27 II) or the nominative uxor flens ‘wife weeping’ (Figure 5.27 III). Brassai uses the images to explain the structural difference between these possibilities.
(5.15) | a. | uxor amans flentem flens acrius ipsa tenebat imbre per indignas usque cadente genas. (Brassai 1873: 7, citing Ovid Tristia) |
b. | My loving wife was in my arms as I wept, herself weeping more bitterly, tears raining constantly over her innocent cheeks. Translation by A. L. Wheeler. Revised by G. P. Goold for the Loeb Classical Library. Available online at https://www.loebclassics.com/view/ovid-tristia/1924/pb_LCL151.21.xml. |
[5.62] Brassai uses a similar graphic later in his book (Brassai 1873: 37), also in a footnote. There he calls his graphic a schemaja. It is unclear where the inspiration for this graphical display originates. Imrényi & Vladár (2020: 166-169) propose that it is Brassai’s own interpretation of Priscian, with possibly an influence from Billroth (cf. Section 3.2.7). Particularly noteworthy is the verb-centric display, which is not previously attested. Brassai appears to be the first to use such an analysis. The next author to use rather similar looking graphical displays is Kern (cf. Section 3.3.24), but there is no evidence that Kern knew about Brassai’s graphical analysis.
[5.63] Albert W. de Groot (1892-1963)
[5.64] brackets (Groot 1949a: 10), numbers and incidental brackets (Groot 1949b: 117, 156, 184, 230)
[5.65] basic underline-style diagramming in (Paardekooper 1955a), published version identical to thesis (Paardekooper 1955b). More extensive diagramming in (Paardekooper 1963)
[5.66] l’abbé Louis Gaultier (1746-1818)(Gaultier 1817) parsing tables, but without numbers
[5.67] (Gaultier 1819) using colors and circles to teach word classes to small kids
[5.68] (Tesnière 1934a), see also (Mazziotta 2019), from “Comment construire une syntaxe”, compare “head in the middle” from petrov
[5.69] Tesniere knew about Fry (Fry 1925), documented in (Mazziotta & Kahane 2024), uses numbers like Sicard
[5.70] Petit grammaire russe with trees more like Kern, look very modern (Tesnière 1934b: 162)
[5.71] precursor of Elements: (Tesnière 1953)
[5.72] maurice bologne (1900-1984), cited by Jespersen (Jespersen 1937: 87-88) L’analyse grammaticale à l’aide de signes conventionnels, originally published 1935 with Georges Thone in Liège. Reissued version with inspiration from Jespersen (Bologne 1963)
[5.73] https://nl.wikipedia.org/wiki/Maurice_Bologne
(5.16) | French (Jespersen 1937: 65; cited by Bologne 1963: 17) | |||||||
L’ | homme | qu’ | il | a | tué | était | Jaures | |
the | man | that | he | have | killed | was | Jaures | |
‘The man whom he killed was Jaurès.’ |
[5.74] first edition not available in Germany? 9th edition probably same as 1950 edition (Bonnard 1970)
[5.75] Original from 1939, but not diagrams. Takes over Bonnard-style diagrams in 25th edition (1959)
[5.76] 26th edtion 1963 https://archive.org/details/precisdegrammair0000grev 28th edition 1969 in Marburg library 30th edition 1995 https://archive.org/details/precisdegrammair00grev
[5.77] I thank María José García Folgado for many suggestions
[5.78] Montoy, Joaquim (1882, mayo 12). “Análisis gramatical razonado”. El Clamor del Magisterio, año 17, n°. 19
[5.79] Tomás Escriche y Mieg (1844-1918) and Francisco Fernández Iparraguirre (1852-1889)
[5.80] large fold-out page between p124-125 (word classes) and between p146-147 (syntactic functions)
Es muy práctico y útil para el conocimiento de esta parte de la Gramática, que los alumnos se acostumbren a ordenar, en la forma que hemos dicho, los diversos elementos de la proposicion; este trabajo no es nuevo, es precisamente el que resulta de practicar el análisis de aquélla por medio del cuadro de la página 146, que nos hace ver desde luego todas las inversiones que existen. (Escriche y Mieg & Fernández Iparraguirre 1884: 153)
[5.81] Francisco Febres Cordero (1854-1910), also known as “brother Miguel” in the Ecuadorian brotherhood Hermanos de las Escuelas Christiana (Montoro del Arco 2023: 169-171).
[5.82] (Cordero 1907a; Cordero 1907b)
[5.83] Cordero was sent to europe, and probably this version was published under his auspices in Paris (Montoro del Arco 2023: 181). Earlier versions from Ecuador do not have any graphical analyses.
[5.84] “análysis lógico” (Cordero 1907b: 28-36)
[5.85] A different level of granularity for the analysis of the sentence is the quadro de subordination ‘subordination scheme’ (explained in Cordero 1907a: 28; more examples in Cordero 1907b: 78, 92, 97, 105)
[5.86] (Solana 1911), nicer pictures in 5th edition (Solana 1911)
[5.87] Bastianini, René (1916). “Análisis sintáctico inmediato y secundario”. Sintaxis castellana y nociones de lingüística y etimología. Buenos Aires: Librería de A. García Santos, 48-49.
[5.88] René Bastianini (1877-1953) (Bastianini 1916: 42-49), see also (García & Folgado 2017)
[5.89] Bouilly, Enrique. (1918). “Bosquejo de un nuevo método de análisis”. El Monitor de la Educación Común. Año 36, n°. 550, p. 39.
[5.90] http://www.bnm.me.gov.ar/giga1/monitor/monitor/550.pdf p.32 https://amsacta.unibo.it/id/eprint/7830/1/129-151%20Esteban%20Lidgett%20et%20al.pdf
[5.91] Montoro del Arco, Esteban T. (2019). “Métodos y modelos de representación del análisis sintáctico en la gramática escolar española (siglos XIX y XX)”, en: Zamorano Aguilar, A. (coord. y ed.) / González-Fernández, A., Rodríguez-Tapia, S. y González Jiménez, J. M. (eds.): Historiografía de la reflexión sintáctica: metaanálisis y estudios en torno al español, München: LINCOM, 303-335. [Colección: Studies in Spanish Linguistics]
[5.92] http://hdl.handle.net/10481/67999
[5.93] https://digibug.ugr.es/bitstream/handle/10481/67999/Metodos_y_modelos_de_representacion_del%20%281%29.pdf?sequence=1
[5.94] Hariton Tiktin (1850-1936) was born in Breslau (Wrocław), but moved to Iași in Romania at the age of 18, where he worked as a teacher. He wrote a dissertation in Leipzig and worked as a lecturer at the Humboldt-University Berlin
[5.95] (Tiktin 1895[1891]: 206-207) second edition, but difference to first edition (1891) unknown
[5.96] Figures look very much like Kern, but no attribution
[5.97] Jan Gebauer (1838-1907) (Gebauer 1900: 237, 249, 257, 265) hierarchical bracketing
[5.98] 1890: https://books.google.de/books?id=QaM3AQAAMAAJ p. 26, 38, 46, 186?
[5.99] ČECHOVÁ, M. Místo tzv. školských mluvnic J. Gebauera v jeho tvorbě a ve vývoji vyučování češtině. In Acta Universitatis Carolinae – Philologica 1–2, Slavica Pragensia XXXV, 1992, s. 79–91.
[5.100] 1894: Historicka Mluvnice Jazyka Ceskeho https://archive.org/details/HistorickaMluvniceJazykaCeskeho.DilI.Hlaskoslovi.JanGebauer./page/n9/mode/2up
[5.101] 1914: Mluvnice česká pro školy střední a ústavy učitelské https://ndk.cz/uuid/uuid:d2a2c090-c25f-11e3-bb44-5ef3fc9bb22f no graphics?
[5.102] (Panevová 2024)
[5.103] Panevová, J. 2016. Syntax Vladimíra Šmilauera včera a dnes. Jazykovědné aktuality 53 (1 a 2): 30–35. https://www.jazykovednesdruzeni.cz/wp-content/uploads/2020/04/2016_1-2.pdf
[5.104] grafické (Novotná 2012; Novotná 2006; Vernerová 2019)
[5.105] Digitalisations:
[5.106] not accessible?
[5.107] Tadeusz Lehr-Spławiński (1891-1965) Roman Kubiński (1886-1957)
[5.108] https://gramatyki.uw.edu.pl/book/544?lang=en
[5.109] first edition 1927, cited here in 2nd edition (Lehr-Spławiński & Kubiński 1928: 150) single tree-image of sentence.
(5.17) | a. | Dziś rano przez dwie godziny padał bezustannie bardzo gęsty śnieg. |
b. | This morning it snowed very heavily for two hours straight. |
[5.110] Zenon Klemenciewicz (1891-1969) structure of complex sentences consisting of multiple clauses (Klemensiewicz 1935: 66-67)
(5.18) | a. | (1) Przed burza bywa chwila cicha i ponura, (2) kiedy, (3) nad glowy ludzi przyleciawszy, (2 c.d.) chmura stanie (4) i (5) grozac twarza, (4 c.d.) dech wiatrów zatrzyma, (6) milezy, (7) obiega ziemie blyskawic oczyma, (8) znaczac te miejsca, (9) gdzie wnet cisnie grom po gromie. |
b. | (1) Before a storm there is a quiet and gloomy moment, (2) when, (3) flying over people’s heads, (2 c.d.) a cloud stands (4) and (5) threatening with its face, (4 c.d.) stopping the winds’ breath, (6) passing by, (7) flashing lightning with its eyes, (8) marking those places, (9) where it will immediately hurl thunder after thunder |
[5.111] https://gramatyki.uw.edu.pl/book/506
[5.112] Encyklopedia języka polskiego (ed. by Urbańczyk, 2nd ed. 1991) calls it wykres składniowy (syntactic graph) and says that as a didactic convention it was popularised (scil. in Poland) by Klemensiewicz. One ex. can indeed be found in his Zarys składni polskiej (Outline of Pol. syntax) of 1953, p. 62.
[5.113] Various textbooks by Klemenciewicz in the 1930s, but no graphics:
[5.114] Piotr Bąk (1911-2000) dialectologist, teacher and later lecturer at the university of Łódź
[5.115] Gramatyka języka polskiego: zarys popularny ‘Polish grammar: a popular outline’ (Bąk 1977) (sentence 400-402, complex sentences: 428-430, 435-442, generative: 446-450)
[5.116] only major revision in 4th edition (Bąk 1984) (sentence 442, complex sentences: 459, 470-487, generative: 497-502) (the illustrations are from 4th ed. of 1984; 1st ed. in 1977; 15th ed. in 2016).
[5.117] https://pl.wikipedia.org/wiki/Piotr_Bąk_(językoznawca)
[5.118] More on old polish grammars, but apparaently no other syntactic graphics: https://gramatyki-uw-edu-pl.translate.goog/?_x_tr_sl=auto&_x_tr_tl=en&_x_tr_hl=de&_x_tr_pto=wapp
[5.119] Info from Tomasz (Ma)ksymil(ian) Majtczak: https://bsky.app/profile/ksymil.bsky.social
[5.120] Fedor Ivanovich Buslaev (1818-1897) was professor for Russian linguistics in Moscow. Some basic biographical information is available on Wikipedia https://en.wikipedia.org/wiki/Fyodor_Buslaev, accessed 20 February 2025. He was strongly influenced by the German linguistic tradition and cites broadly throughout the German literature. Although he does not explicity cite any influence for his graphical approach, he might have been inspired by Diesterweg (see Section 3.2.6) and Mager (see Section 3.3.6), both of which are cited recurrently by Buslaev.
[5.121] In his book О преподавании отечественного языка (‘On the teaching of the native language’) he discusses the analysis of large sentences with multiple subordinate clauses (Buslaev 1844: 260). As an example he give an analysis of a fable (5.19) by using a graphical structure with large horizontal braces, as shown in Figure 5.32. The original structure of this sentence has a lot of clausal centre-embedding. This is indicated in the graphical analysis by the braces, indicating the insertion point of the subordinate clause.
(5.19) | Krylov’s fable The wolf and the shepherds, cited in Buslaev (1844: 260) |
[5.122] Karl Yakimovich Lugebil (1830-1887) was a classical scholar, teacher and from 1864 professor in St. Petersburg. Some basic biographical information is available online at https://www.biografija.ru/biography/ljugebil-karl-yakimovich.htm, accessed 22 February 2025. In a review of a book by Anton Dobiasz about the Syntax of Apollonius Dyscolus he uses various graphical syntactic analyses (Lugebil 1883: 127-128). In his review, Lugebil mostly cites German grammatical literature, for example there is a long discussion about the merits of Becker’s grammatical ideas (cf. Section 3.2.5). Later, in a discussion about different interpretations of copula constructions, Lugebil cites Lehmann (see Section 3.2.8) and Nägelsbach (see Section 5.1.1) as important proponents of the idea to use graphical representations to explain grammatical structure:
По этому поводу обращаю вниманіе гг. преподавателей на то, какь важнодля нагляднаго объясненія-прибегать кь графическому изображенію какь отношеній различныхь составныхь частей предложенія другь кь другу, такь и отношеній подчиненія ихь другь другу или сочетанія различныхь предложеній, входящихь вь составь періодовь античной речи. (Lugebil 1883: 128, fn. 1)
(‘In this regard, I would like to draw the attention of Messrs. teachers to the importance of resorting to a graphic representation for a visual explanation of both the relationships of the various constituent parts of a sentence to each other, as well as the subordinate relationships to each other or the combination of various sentences that make up a period in classical language.’)
[5.123] Lugebil uses a completely different graphical format from Lehmann and Nägelsbach, illustrated here in Figure 5.33. Crucially, Lugebil uses a system of hierarchical bracketing, which does not appear at this point in time in the German tradition. Lugebil’s graphics are used to illustrate two different interpretations of the Latin sentence (5.20). Although this sentence clearly means ‘Cicero became consul in that year’, Lugebil argues that there is an “older” (прежній) interpretation, roughly equivalent to ‘Cicero is a consul, installed in that year’. This “older” interpretation is shown to the left in Figure 5.33, while the regular “newer” (позднійшій) interpretation is shown to the right.
(5.20) | Latin (Lugebil 1883: 128) | |||||
Cicero | illo | anno | consul | factus | est | |
Cicero | dem.abl | year.abl | consul | make.part | cop.3sg | |
‘Cicero became (lit. was made) consul in that year.’ |
[5.124] Konstantin Fedorovich Petrov (1849-1914)
[5.125] 1881 syntax version https://vivaldi.dspl.ru/bx0005325 https://www.prlib.ru/en/node/363131 1886 etymology version https://vivaldi.dspl.ru/bx0005310
[5.126] first edition? Version from 1898 (“Syntax”) does not have graphics: https://books.google.de/books?id=h1M5AQAAMAAJ&pg=PA122
[5.127] 1912 (20th edtion, syntax): graphs on p.127 https://google.com/books?id=qsJTzwEACAAJ better edition: https://google.com/books?id=08gzAQAAMAAJ
[5.128] 1913 (24th edition, etymology): 77 https://rusneb.ru/catalog/000202_000006_2570896/
[5.129] 1915 (27th edtion): 77 https://imwerden.de/publ-1047 “PRACTICAL TEXTBOOK OF RUSSIAN GRAMMAR. ETYMOLOGY”
[5.130] 1923: 167 (31st edition) with graphic https://www.prlib.ru/item/363130
[5.131] https://www.prlib.ru/en/node/363130#i=5640518_doc1_A22FE1CB-0AC5-435A-AF11-A60467AC2B73.tiff
[5.132] 1923: 115 version (32nd edition) does have graphic http://e-heritage.ru/Book/10075936, reprint 1928 as 32nd edtion
[5.133] (Thümmel 1993: 171) cites a version from 1906: 127
[5.134] Ušakov, Dmitrij Nikolajevič https://www.prlib.ru/en/history/618974 Smirnova, A. M Ščepetova, Nina Nikolajevna
[5.135] volume 2, 15th edition: (Ušakov, Smirnova & Ščepetova 1930: 25-28)
[5.136] => 3rd part, pp. 6–7, and 4th part, p. 5
[5.137] nothing in https://www.prlib.ru/en/node/341406
[5.138] Admoni: Partiturnoje strojenije recevoj cepi i sistema grammaticeskich znacenij ν predlo/.enii [Die „Partitur”-Struktur der Redekette und das System grammatischer Bedeutungen im Satz]. In: Naucnyje doklady vysäej äkoly. Filologiceskije nauki [Wissenschaftliche Beiträge der er Hochschule. Philologische Wissenschaften]. 1961, H° 3. - 253, 272, 278, 288
[5.139] see also 4th edition 1982, p. 311-317 https://archive.org/details/derdeutschesprac0000admo
[5.140] Chinese version of reed/kellog: Li Jinxi New Grammar of the National Language (Mossner 1960; Peverelli 2015: 97-104)