A Forgotten Source in the History of Linguistics: Husserl’s Logical Investigations

Attached document(s)

Annexes

• Texte complet en PDF/Full text in PDF: Vol. XI, n°5 (254k)

Abstract

In appearance, Husserl’s writings seem not to have had any influence on linguistic research, nor does what the German philosopher wrote about language seem to be worth a place in the history of linguistics. The purpose of the paper is exactly to contrast this view, by reassessing both the position and the role of Husserl’s early masterpiece — the Logical Investigations — within the history of linguistics. To this end, I will focus mainly on the third (On the theory of wholes and parts) and fourth (The distinction between independent and non-independent meanings) Investigations, paying special attention to Husserl’s mereology and to the idea of a general pure grammar. The paper tries to situate the third and fourth Logical Investigation within the general context of late nineteenth-century and early twentieth-century linguistics and furthermore attempts to show the historical and theoretical importance of the Logical Investigations for the birth and the development of one of the most important linguistic “schools” of the twentieth century, namely structural linguistics.

Table of content

1. Husserl and the History of Linguistics

1In most histories of linguistics, it is usual to read that a nineteenth-century characterized by historical-comparative linguistics gives the way to a twentieth century marked by theoretical or general linguistics. The first expression, historical-comparative linguistics, indicates the branch of linguistics which is concerned with the scientific study of language change over time, having as its principal goals the account of observed changes in particular languages, the reconstruction of the pre-history of languages, the grouping of languages into language families on the basis of their supposed relatedness, the development of general theories about language change, the study of the history of words and of speech communities with the objective of gaining a better understanding of the phenomenon of linguistic change.1 General or theoretical linguistics, on the other hand, is mostly interested in establishing “general principles for the study of all languages, and to determine the characteristics of human language as a phenomenon”,2 without being restricted to a particular language. It therefore involves the search of linguistic universals, that is of properties which are common to all human languages.

2Although the mentioned periodization represents a quite rough simplification — since theoretical and general approaches are in truth traceable also in nineteenth-century linguistics, most blatantly in Wilhelm von Humboldt’s inquiries, and historical linguistic studies were constantly growing also in the last century — it is nonetheless based on an indisputable historical fact: indeed, historical linguistics arises at the beginning of the nineteenth century with scholars like Bopp, Rask and Grimm and definitively defines its techniques and principles at the end of the century with the work of the so-called neo-grammarian school, of which the German linguist Hermann Paul, mentioned by Husserl in the fourth investigation, can be considered a prominent exponent. On the other hand, as John Lyons notes in his Introduction to Theoretical Linguistics, already “[…] by the end of the eighteenth century there had developed a general dissatisfaction with a priori and so-called ‘logical’ explanations and a preference for historical reasoning”.3 As a matter of fact, theoretical approaches were indeed less frequent in the nineteenth century, although present, and anyway not able to achieve systematic results and to develop a set of established common techniques until the early twentieth century and, more precisely, until the scientific work of the Swiss linguist Ferdinand de Saussure and the consequent publication of the text which has been considered as his masterpiece, namely the Cours de linguistique générale, which dates from 1916 and which opens the way to what has significantly been called “the Saussurean break”.4

3What is crucial in the Saussurean break is the introduction of a systemic approach. Indeed, as Émile Benveniste — one of the most influential linguists of the twentieth century — writes, “la nouveauté de sa doctrine est là, dans cette idée, riche d’implications qu’on mit longtemps à discerner et a développer, que la langue forme un système”.5 In fact, as stated by Saussure, “la langue est un système dont tous les termes sont solidaires et où valeur de l’un ne résulte que de la présence simultanée des autres”.6 “La langue”, Saussure writes in the third chapter of the first part of the book, “est un système dont toutes les parties peuvent et doivent être considérées dans leur solidarité synchronique”.7 According to this view, linguistics must therefore be defined as the science of language understood as a system of interdependent elements and its objective must furthermore be in the first place, with Saussure’s words, “de chercher les forces qui sont en jeu d’un manière permanente et universelle dans toutes les langues, et de dégager les lois générales auxquelles on peut ramener tous les phénomènes particulier de l’histoire”.8 Thus, although in his 1916 masterpiece Saussure uses the word “structure” very rarely — and never in a technical sense — his linguistic theory nonetheless can be said to pave the way for twentieth-century theoretical approaches to language and, more specifically, to structural linguistics. In fact, in structural linguistics — as Roman Jakobson, a prominent spokesperson for structural linguistics, writes in a 1958 essay — “[one] speak[s] about the grammatical and phonological system of language, about the laws of its structure, the interdependence of its parts, and of the parts and the whole”.9 As Émile Beveniste summarizes,

[i]l s’agit donc, la langue étant posée comme système, d’en analyser la structure. Chaque système, étant formé d’unités qui se conditionnent mutuellement, se distingue des autres systèmes par l’agencement interne de ces unités, agencement qui en constitue la structure. Certaines combinaisons sont fréquentes, d’autres plus rares, d’autres enfin, théoriquement possibles, ne se réalisent jamais. Envisager la langue (ou chaque partie d’une langue, phonétique, morphologie, etc.) comme un système organisé par une structure à déceler et à décrire, c’est adopter le point de vue “structuraliste”.10

4As is well known, Husserl was among the first to develop, in the third Logical Investigation, a formal mereology, that is a formal study of the relations between parts and whole as described by Jakobson. A prominent contemporary philosopher like Kit Fine even wrote that “Husserl’s third Logical Investigation is perhaps the most significant treatise on the concept of part to be found in the philosophical literature”.11 Furthermore, in the Fourth Logical Investigation Husserl applies the results achieved in the third investigation to the field of meaning and language and aims to describe a set of formal laws governing the possible combinations between expressions — and hence, as we shall see, between meanings — of the type referred to by Benveniste. Moreover, Husserl makes explicit references to the history of linguistics and to contemporary linguists like, for instance, Heymann Steinthal, who was the teacher of many representatives of the neo-grammarian school at the University of Berlin.

5Given the scientific framework that dominated linguistic studies in the years in which he writes his Logical Investigations — namely the years between 1894 and 1901 — Husserl’s stance on language turns out to be, then, neither expected nor ordinary. In fact, he sharply criticizes mainstream historical-comparative approaches to the study of language12 and instead considers his position as a revival and a development of, using Husserl’s words, “die alte Idee einer allgemeinen und sogar einer apriorischen Grammatik”.13 The allusion, as is made explicit in the last pages of the fourth investigation, is to the Grammaire générale et raisonnée, otherwise known as the Port-Royal Grammar, developed by Antoine Arnauld and Claude Lancelot in the second half of the seventeenth century. Without going into further details, the Port-Royal Grammar can be said to be general, to the extent that is not concerned with specific languages but, on the contrary, aims to discover the universal features of human language in general and is furthermore reasoned insofar as it connects the functioning of language to the functioning of reason, namely to logic.14

6Moreover, Husserl’s stance towards the linguistic studies of his time is not only unusual but also completely aware, as shown by the following quotation taken from the fourth Logical Investigation:

Nun ist in unserem naturwissenschaftlichen Zeitalter dafür gesorgt, daß empirisch-allgemeine Forschungen, wie überall, so in grammatischen Dingen, nicht unterlassen werden. Anders steht es mit den apriorischen, für die der Sinn in unserer Epoche fast zu verkümmern drohte, obschon doch alle prinzipiellen Einsichten auf sie zurückführen.15

2. Mereology

7Husserl’s idea of an a priori or pure grammar, which can be seen both as a revival of the general conception of the Grammaire générale et raisonnée and as anticipation of twentieth-century structural linguistics, is based on the mereology developed by him in the third investigation. In this investigation, however, Husserl combines the mereological framework with an ontological commitment, insomuch as it is possible to consider this investigation as a study providing the outlines of a formal ontology, that is of an ontology whose concepts, as Barry and David Smith write, “are independent of any peculiar material of knowledge”.16 Given this ontological commitment, the mereology proposed by Husserl implies that the analysed relations between parts and wholes must be considered as relations between objects and set of objects, where the term “object” has to be understood in its broadest sense and precisely, according to the early Husserl, as simply indicating a representative non-contradictory content. Hence, a bottle, a feeling, a sound or an angel can all be described as objects while, for instance, a “square round” or a “wooden iron”, since contradictory, cannot.

8The first axiom, as we could also call it, of Husserl’s formal ontology says the following: “Jeder Gegenstand ist wirklicher oder möglicher Teil, d. h. es gibt wirkliche oder mögliche Ganze, die ihn einschließen”.17 This means that every object is always and necessarily part of a whole. In other words, it is impossible for an object to be absolutely and completely self-sufficient, since there is always a whole, real or possible, which contains it or which might contain it. Every object is then necessarily a part, it is then always included in a web of relations, which connects it with other objects. Hence, in this sense, all objects must be said to be, using Husserl’s wording, non-independent. Still, there is another sense according to which objects, or parts, can instead be said to be relatively independent or relatively non-independent. Husserl writes,

[e]in Inhalt x ist relativ unselbständig zu einem Inhalt ß, bzw. zu dem durch ß und alle seine Teile bestimmten Gesamtinbegriff von Inhalten , wenn ein [...] reines Gesetz besteht, wonach überhaupt ein Inhalt der reinen Gattung x a priori nur in oder verknüpft mit anderen Inhalten aus dem durch ß bestimmten Gesamtinbegriff von reinen Inhaltsgattungen bestehen kann. Mangelt ein solches Gesetz, so nennen wir x relativ zu ß selbständig.18

9Thus, if an object “A” is connected with an object “B” through a law of necessary implication, this means that “A” is relatively non-independent towards “B”, while if the connection between “A” and “B” is of an arbitrary or accidental nature, one can say that “A” is relatively independent towards “B”. On this basis, Husserl then formulates, in the third investigation, a number of laws concerning the kind of relation represented by necessary implication, for example the law according to which the existence of a whole “G” composed by the elements “A”, “B”, “C”, necessarily implies the existence of the elements “A”, “B” and “C”.

10Husserl calls the relation of necessary implication between two objects a relation of “foundation”, therefore we can say that an object “A” is relatively non-independent towards “B” if “A” is founded in “B” and, on the contrary, relatively independent if “A” is not founded in “B”. The kind of necessity Husserl speaks of, is not, as he writes, “[eine] subjektive Notwendigkeit, d. i. subjektive Unfähigkeit des Sich-nicht-anders-vorstellen-könnens, sondern objektiv-ideale Notwendigkeit des Nicht-anders-sein-könnens”.19 The kind of necessity that comes into play in the relations of independence and non-independence is namely an ontological necessity and not a mere psychological one.20 In a relation of independence, then, what prevents the being otherwise of something is, according to Husserl,

das Gesetz, das sagt, es ist nicht bloß hier und jetzt so, sondern überhaupt, in gesetzlicher Allgemeinheit. Nun ist aber zu beachten, daß, wie die Notwendigkeit, von der hier, in unserer Erörterung der „unselbständigen“ Momente, die Rede ist, die Bedeutung einer in den sachlichen Wesen gründenden idealen oder apriorischen Notwendigkeit hat, so korrelativ die Gesetzlichkeit die Bedeutung einer Wesensgesetzlichkeit, also einer nicht-empirischen, unbedingt allgemeingültigen Gesetzlichkeit.21

11On the basis of these analyses, Husserl can then distinguish two kinds of sets, namely “aggregates” (Inbegriffe) and “wholes” (Ganzen). Aggregates are sets of objects, which stand together accidentally, that is without implying a relation of foundation, whereas by wholes Husserl understands

einen Inbegriff von Inhalten, welche durch eine einheitliche Fundierung, und zwar ohne Sukkurs weiterer Inhalte umspannt werden. Die Inhalte eines solchen Inbegriffs nennen wir Teile. Die Rede von der Einheitlichkeit der Fundierung soll besagen, daß jeder Inhalt mit jedem, sei es direkt oder indirekt, durch Fundierung zusammenhängt.22

12I believe that the notion of “whole” (Ganzen) proposed by Husserl in the third investigation is perfectly comparable to the structuralist notion of “structure”. After all, as Göran Sonesson has noted in a recent article on structuralism, “structure has to be studied within a more complete mereological framework, that is, within the science of parts and their relation to the whole, first defined by Twardowski and Husserl”.23Moreover, just in the same way as for the notion of structure, a whole is not merely the resulting sum of its components. When one perceives a complex object like, for instance, a table, one does not simply perceive the sum of its components but rather the table as a unitary whole, as a structure, that is primarily as a homogeneous arrangement composed by a set of elements and of interconnected relations, a set which can then be broken into different parts like, in the case of the table, its legs and surface and into different relations like, for instance, the “being-beneath” of the legs of the table in regard to its surface.

3. Pure Grammar

13In the fourth investigation, Husserl applies this theoretical framework to a very special domain of objects, namely to expressions, which, in the terminology of the Logical Investigations, means linguistic signs which bear a meaning, that is a reference to a class of objects.24 Like every kind of objects, expressions are also either independent — and in this case they bear an independent meaning — or non-independent — and in this case they bear a non-independent meaning. Since in the fourth investigation Husserl is primarily interested in the possible relations between expressions, and therefore between meanings, he mainly focuses his attention on sets of expressions, namely on what he calls complex expressions like, for instance, the sentence “a king who wins the love of his subjects”. Husserl then distinguishes between what he calls, following the terminology proposed by the philosopher and linguist Anton Marty, syncategorematic expressions like, in the exemplifying sentence, “a”, “who”, “the”, “of”, “his”, and categorematic expressions such as, in this case, “king”, “wins”, “love”, “subjects”.

14Syncategoremata are, according to Husserl, “Zeichen […] welche nur mit anderen Redebestandteilen zusammen eine vollständige Bedeutung haben”,25 while categorematica are “für sich bedeutsamen oder vollständigen Ausdrücke”.26 In fact, the syncategorematic expression “who” has no meaning outside the complex expression in which it occurs, while the categorematic expression “king” maintains his meaning even if isolated from the broader expression in which it occurs. Thus, meanings borne by syncategorematic expressions are always non-independent, while meanings borne by categorematic expressions are always independent. This means that non-independent meanings are, following the terminology introduced by Husserl in the third investigation, always founded (fundiert) in other meanings, that is to say that they necessarily imply other meanings in order to be, in turn, meaningful themselves. Indeed, while the isolated expression “king” can be said to have a complete meaning for itself, the isolated expression “who” has no meaning unless connected with other expressions like, for instance, in the sentence “the king, who…”.

15“Zu jedem Fall einer unselbständigen Bedeutung”, Husserl then writes, “gehört […] ein gewisses Wesensgesetz, welches ihre Ergänzungsbedürftigkeit durch neue Bedeutungen regelt, also die Arten und Formen von Zusammenhängen nachweist, in denen sie eingeordnet sein muß”.27 The study and identification of these laws of essence concerning the possible combinations of expressions within language is exactly the concern of what Husserl calls “pure grammar”. Indeed, according to Husserl, “[e]s gibt im Bedeutungsgebiete apriorische Gesetzmäßigkeiten, wonach Bedeutungen bei Erhaltung eines wesentlichen Kerns [...] in neue Bedeutungen umzuwandeln sind”.28 Take, for instance, the complex expression “the tree is green”. Of course one can vary the form of this expression in many different ways producing new expressions like “the tree is beautiful”, “the kid is green”, “the green is beautiful”, but not every variation turns out to be in truth possible. Indeed, expressions like “the kid is the.”, “the green is or.”, “the or is the.” cannot be considered as possible combinations. According to Husserl, the impossibility of these combinations

ist eine wesensgesetzliche, d. h. zunächst, sie ist keine bloß subjektive, es liegt nicht bloß an unserer faktischen Unfähigkeit (an dem Zwange unserer “geistigen Organisation”), daß wir die Einheit nicht vollziehen können. In den Fällen, die wir hier im Auge haben, ist die Unmöglichkeit vielmehr eine objektive, ideale, in der “Natur”, im reinen Wesen des Bedeutungsgebietes gründende und als solche durch apodiktische Evidenz zu erfassen.29

16What Husserl means, is that the kind of possibility and impossibility involved in the considered combinations of expressions does not limit itself to a specific empirical language or to a particular individual psychological speaker but rather concerns universal foundational laws which determine the semantic and syntactic organization of language as such.30 In fact, according to Husserl,

die Sprache nicht bloß ihre physiologischen, psychologischen und kulturhistorischen, sondern auch ihre apriorischen Fundamente hat. Letzteres betrifft die wesentlichen Bedeutungsformen und die apriorischen Gesetze ihrer Komplexion bzw. Modifikation, und keine Sprache ist denkbar, die durch dieses Apriori nicht wesentlich mitbestimmt wäre […] Wieviel vom tatsächlichen Inhalt der historischen Sprachen sowie von ihren grammatischen Formen in dieser Weise empirisch bestimmt sein mag, an dieses ideale Gerüst ist jede gebunden; und so muß die theoretische Erforschung desselben einesder Fundamente für die letzte wissenschaftliche Klärung aller Sprache überhaupt ausmachen […] Man muß es vor Augen haben, um sinnvoll fragen zu können: Wie drückt das Deutsche, das Lateinische, Chinesische usw. „den“ Existenzialsatz, „den“ kategorischen Satz, „den“ hypothetischen Vordersatz, „den“ Plural, „die“ Modalitäten des „möglich“ und „wahrscheinlich“, das „nicht“ usw. aus?31

17As Husserl writes in the passage just quoted, there are, according to him, two main classes of essential laws concerning the field of expressions, that is of meaningful signs, namely laws of compounding (Komplexion) and laws of modification (Modifikation). While the first ones establish the possibility or impossibility of certain combinations between meanings or, better, between meaning categories,32 the second ones govern the way in which the regular functioning of the first ones can be altered by means of a transformation of a meaning category into another one, as in the cases of nominalization, in which verbs, adjectives, adverbs, etc. are used as nouns.

18An example of what Husserl understands by a law of compounding, is represented by the following rule: “[es] gehört zu je zwei nominalen Bedeutungen M und N die primitive Verknüpfungsform M und N, mit dem Gesetze, daß das Verknüpfungsresultat wieder eine Bedeutung derselben Kategorie ist”.33 With reference instead to the laws of modification, Husserl gives, among others,34 the following example:

Jeder Ausdruck, gleichgültig ob er — in seiner normalen Bedeutung — ein kategorematischer oder synkategorematischer ist, kann danach als Name von sich selbst auftreten, d. h. er nennt sich selbst als grammatische Erscheinung […] Sagen wir „und“ ist eine Konjunktion, so haben wir nicht das Bedeutungsmoment, das dem Worte und normalerweise entspricht, an die Subjektstelle gebracht, sondern hier steht die selbständige, auf das Wort und gerichtete Bedeutung. In dieser anomalen Bedeutung ist das und in Wahrheit kein synkategorematischer, sondern ein kategorematischer Ausdruck, es nennt sich selbst als Wort.35

19On the basis of these analyses, Husserl distinguishes two kinds of possible incompatibilities. On the one hand, he designates with the term Nonsense (Unsinn) all the incompatibilities deriving from the violation of the laws of compounding, whereas, on the other hand, he labels with the term Absurdity (Widersinn) a special class of complex meanings, namely the class of complex meanings whose reference objects are contradictory.

20In the first case, we are faced with no more than aggregates of meanings, with mere sums of meanings, which stand together in a totally arbitrary and accidentally way, without implying any form of necessary interaction and which therefore cannot be described as wholes. The examples made by Husserl on this regard include sentences like “A round or.”, “A man and is.”, etc.36

21In the second case, instead, we deal with complex meanings which do not violate any law of compounding and which are not, accordingly, mere aggregates but can, on the contrary, be considered as proper wholes. However, since the objects these complex meanings refer to are contradictory, the meanings forming the whole can still be said to be, to some degree, incompatible with each other, insofar as they hinder the normal functioning of the process of reference. Typical examples of absurdity are then, according to Husserl, expressions like “the round square” or «the wooden iron”.

4. Conclusion: Husserl and Structural Linguistics

22On the basis of the line of reasoning outlined above, it is now possible to highlight some substantial theoretical and historical affinities between structural linguistics and early phenomenology, to the extent that is even possible to consider the Logical Investigations as one of the fundamental sources of linguistic structuralism.37 Indeed, although it is true that there seem to be no direct contacts between Saussure’s and Husserl’s works, it is on the other hand true that there are many meeting points, both at historical and theoretical levels, between Husserl’s early philosophy and the scientific activity of Roman Jakobson and, more generally, of the other members of the Prague linguistic circle, founded in 1926 by the Czech linguist Vilem Mathesius and considered, together with the Copenhagen school, the place in which structural linguistics took its most rigorous form.38 Jakobson is certainly the scholar who brings the claims of early phenomenology — and above all of the Logical Investigation39— most clearly into the field of structural linguistics. Indeed, Jakobson explicitly considers Husserl’s phenomenology as one of the main sources underlying Prague structuralism and Russian formalism and, more specifically, defines Husserl’s third Logical Investigation as “[eine] strukturalistische Fundamentalbetrachtung”.40 It is significant to notice that, in some cases, Jakobson even uses, as Holenstein observes, “[die] von Husserl […] inspirierten Formulierung (‘Fundierung’ statt ‘Implikation’)”.41Furthermore, it is noteworthy to mention that on the 18^th of November 1935, at the invitation of Jakobson himself, Husserl gave a lecture, entitled Phänomenologie der Sprache and unfortunately lost, at the Prague linguistic circle.42

23Moreover, without being able to go into further details here, it is important to mention the relevance of the role played by two fundamental, although rather neglected, mediators between Husserl’s phenomenology and Jakobson’s linguistic thought, namely Gustav Špet and Hendrik Pos, to whose works Patrick Flack has interestingly drawn attention recently.43

24Also on the Danish side of structural linguistics, the influence of Husserl is, although less evident, certainly significant. The three kinds of dependence identified by Hjelmslev in the Prolegoma to a Theory of Language, for instance, are closely connected to the relations of foundation (Fundierung) distinguished by Husserl in the third Logical Investigation.44 Moreover, the influence of Husserl becomes explicit as soon as we turn to other members of the Copenhagen school like Viggo Brøndal and Paul Diderichsen, this latter a student of both Hjelmslev and Brøndal. Indeed, while Diderichsen,as reminded by Federik Stjernfelt, “[a]t several occasions, remarks upon the complete similarity between the dependence calculi of the Prolegomena [to a Theory of Language] and Logical Investigations”,45Brøndal, for his part, concludes an important essay of 1939 with the following passage, which is worth quoting in its entirety, given the similarities in the usage of mereological notions in relation to the conceptual framework developed by Husserl in the third and fourth investigation:

On a conçu […] la structure comme objet autonome et par conséquent comme non-dérivable des éléments dont elle n’est ni l’agrégat ni la somme; c’est pourquoi il faut considérer l’étude des systèmes possibles et de leur forme comme étant de la plus grande importance. — Et pourtant on ne saurait considérer les éléments qui font partie d’un système comme de simples dérivés des corrélations ou oppositions structurales; il sera en effet important de distinguer entre les propriétés purement formelles d’un système et sa matière ou substance qui, tout en étant adaptée à la structure donnée (puisqu’elle y entre), n’en est pas moins relativement indépendante; et l’étude des catégories réelles, contenu ou base des systèmes, sera non moins importante que celle de la structure formelle. Les méditations pénétrantes de Husserl sur la phénoménologie seront ici une source d’inspiration pour tout logicien du langage.46

25Finally, from a theoretical point of view, it is possible to sum up — on the basis of the analyses developed in the second and third section of the present paper — the main debts owed by structural linguistics towards Husserl’s early phenomenology in the following points:

26(1) The mereological framework and the concept of whole.

27(2) The predilection of theoretical, general, a priori and formal approaches to the study of language and the consequent criticism of (linguistic) psychologism,47 that is of the view according to which “language has to be understood and explained through psychology”,48 namely through an empirical and experimental science that can provide mere factual, contingent and particular rules but no logically necessary objective truths.

28(3) The Idea of a pure grammar and of a “combinatorics” of meanings and expressions.49

Bibliographie

Notes

To cite this article

About: Simone Aurora