The (early) history of linguistic typology,

THE [OXFORD] HANDBOOK OF LANGUAGE TYPOLOGY (Oxford University Press 2006) Paolo Ramat (Pavia): The (early) history of linguistic typology. Preliminary version: please do not quote 0. It is hard to state when the (early) history of linguistic typological studies started, just as it is difficult to state when the beginnings of, say, neurology or neurolinguistics have to be posited. Broca is a pioneering milestone, but certainly his interests did not arise from scratch. People such as Franz Joseph Gall (1758-1828) and many others had already proposed a cerebral localization of mental faculties. Every science, every discipline knows what can be called an ‘incubation phase’ during which the bases of the scientific problems are laid in general terms before they become the object of scientific research in the proper sense, endowed with a method of its own and fitting its specific research domain. The same holds for linguistic typology. The more so, when compared with the previous example of neurolinguistics, since studying languages according to their structures and systems posits no problems from the ethic or religious point of view; and this is not the case for neurolinguistics. The study of language structures did and does not require to handle corpses and has never been forbidden for religious or ethic reasons, so that it can be said that the study of linguistics (in the broader sense) has started some millennia ago. [space] 2 1. We understand by linguistic typology the systematic cross-linguistic comparison that aims at discovering the underlying universal properties of human language. If we adopt such definition it is clear that the first sections of this chapter may be only marginal to typology in the modern sense of the term. Nevertheless it is interesting to follow the evolution of (western) thought which can be considered as preliminary steps towards typology. Though Greeks and, later, Romans did show a real interest for ‘exotic’ populations (suffice to think of Herodotus’ Historíai), linguistic diversity was in classical antiquity never in the focus of the interest. There are very important theoretical discussions on language, the most relevant being Plato’s Kratylos and Aristotle’s Categories; but the ‘differentia linguarum’ is not in the focus of those discussions. Even the term grammatikḗ from the 3rd century BC. onwards refers to what we would now call ‘philology’ or ‘literary criticism’, including also rhetoric. The most relevant contribution of the grammatikḗ has doubtless been the theory of the ‘parts of speech’ (mérē toû lógou, partes orationis), such as noun, verb, pronoun, adverb, etc. with the features of singular/dual/plural, masc./neuter/femin., present/future/past etc: both the categories and their features have constantly been used in the western typology up to present times. The borrowing of grammatikḗ in the Latin tradition (gram(m)atica) refers both to the teaching of the linguistically correct forms and, again, literary criticism; later on, also to rhetoric (Donatus, 4th cent. AD) and dialectic 3 (Augustin, 4th cent.). From Varro (2nd cent. AD) and Isidor (6th-7th cent.) we know that there was a keen interest on etymology as the search of the ‘true’ (étymos) meaning of the words. But also in this particular field no crosslinguistic comparison was attempted. Given the cultural primacy of the Greek language all over the Eastern part of the Mediterranean, when Rome came in contact with Greece, Roman gramatici simply tried to adapt the categories elaborated in the tradition of the grammatikḗ (Dionysius Thrax 2nd-1st cent. BC, Apollonius Dyskolos 2nd cent. AD, etc.) to their own language (cp. Matthews 1990: 195). The ‘parts of speech’ which Dionysios Thrax had singled out in his grammar (téchnē grammatikḗ) were taken over by the Romans with just some minor changes. Unfortunately the syntactic description of Latin which formed the second part of the De lingua latina written by Varro (106-27 BC) ─probably on the track of Apollonius’ Perì syntákseōs─ has been totally lost. We know very little of the syntactic theories of classical antiquity. The most popular grammar in the Middle Ages, that of Priscian (Mauretania, 6th cent.), who is the author of a voluminous treatise ─Institutio de arte grammatica─ was directly inspired by Apollonius Dyskolos’ Perì syntáxeos. Indeed Ancient Greek and Latin are morphosyntactically very similar, so that there was no real need to develop a contrastive study of the two languages (which is often the basis for a typological approach). [space] 4 2. Reasons of space do not permit to tackle here the important contribution of the Middle Ages to the general discussion on language, namely the logicalsemantic analyses of the 12th-14th centuries by scholars such as Petrus Abelard, Thomas of Aquin, Duns Scotus or Wilhelm of Ockham. In our frame of reference which looks at the attention paid to linguistic differences as the starting point for a possible typological approach the Medieval scholastica does not play an important role, though its role in defining the parts of speech according to the inherited Latin tradition has been on the contrary very important: the grammar of a particular language (Latin) became the basis for a general theory of grammar. Thanks to the so-called modistae such as Martin of Dacia (¶1304) and Michel of Marbais (13th cent.), Thomas of Erfurt (14th cent.), who commented on Priscian’s treatise, much attention was paid to the ways of expression (modi significandi) in the frame of a general, universal Grammatica speculativa, which disregarded how this grammatica is implemented in different languages. The modi significandi are described with reference to the different ‘parts of speech’ of the Latin and Medieval tradition. Things changed when attempts were done to describe new languages according to the Latin grammar. This is the case of the Irish ‘Auraicept na nÉces’, the Icelandic so-called ‘First grammatical Treatise’ (second half of the 12th cent.) and the ‘Donatz Proensals’ (ca. 1240). The Irish and Icelandic phonetic systems with, e.g., the consonantal lenition or, respectively, the vowels posited problems of transcription into the Latin alphabet; whereas the 5 Provençal bicasual system did not match the traditional case system of Latin. Thus the problem of the existence of different linguistic system was implicitly posited (cp. Vineis and Maierù 1990: 85- 92). Dante Alighieri (1265-1321) recognizes that all humans are endowed with the speaking ability, though languages may differ according to different dispositions of their nature. Thus, it is implicitly maintained that language differentiation is not the consequence of the Babel sin, but a natural fact (see below, section 4 on Leibniz). Then Dante considers in his De vulgari eloquentia (1303 ?) three language families in Europe (Germanic, Latin, and Greek) and makes the well-known distinction of the Romance area between the oïl-, oc- and sì-languages according the way these languages say “yes”. We are of course yet very far from linguistic typology, but it is however noteworthy that the first step toward a classification according to linguistic characteristics has been done. [space] 3. The Renaissance is the period wherein the European ‘vulgars’ receive their standardization and consequently the difference among languages becomes a problem to be discussed. A competition arises among scholars of different countries as to which language may boast its primacy. The crosslinguistic comparison is however conducted without the scientific, i.e. falsifiable tools we are accustomed to since the 19th century. Rather, ideological, political and religious considerations are adduced to ‘prove’ the 6 précellence (primacy) of this or that language. The idea of a primitive, pure language before Babel is a key point in many discussions: the language which can be ‘proved’ nearest to this lost original language deserves the précellence. So, for instance, French Humanists maintained that French derived from Hebrew, the language Adam spoke in the lost Paradise and his son Noah brought to Gallia, whilst the Florence humanistic Academy maintained that Tuscan was the heir of Etruscan, which sprang from Aramaic ─again the language spoken by Noah after the Flood. The Flemish Johann Goropius Becanus (1519-1572) is perhaps the clearest example of this ‘linguistic’ treatises where religious belief on the Bible narration and national pride played the prominent role: in his Origines antverpianae he affirms that the dialect of Antwerp is the oldest and hence the purest language of Europe, since the town was founded by the Cimmerii or Cimbri who descended from Japhet’s eldest son, Gomer (called also Cimen/Comen). In the midst of such dissertations some more data based works deserve to be mentioned, such as the Grammaticae quadrilinguis partitiones (1544) by Johannes Drosaeus (Jean Drosée), the first comparative grammar of French, and the three ‘holy languages’, Hebrew, Greek and Latin. Drosaeus makes the comparison on the basis of the traditional Latin grammars (Priscian and Donatus) and show that the tools for the analysis of Latin are valid also for the three other languages. Also the first grammars of English and German were 7 not only written in Latin but used the well-established categories of that language. Be it as it may, it is a historical fact that the cross-linguistic dimension was established in the Renaissance period (Tavoni 1990: 216 f.). [space] 4. The nearer we approach to our times, the more the problems dealt with by scholars such as Francis Bacon, Antoine Arnauld and Claude Lancelot (Grammaire générale et raisonnée: 1660), Thomas Hobbes, John Locke, Gottfried von Leibniz, Giambattista Vico, Etienne Bennot de Condillac, Du Marsais and Beauzée (in Diderot and D’Alembert’s Encyclopédie: 1751-1772), to quote just the most prominent, are relevant to the present linguistic discussion. There exists a large literature on these names, starting with Chomsky’s Cartesian linguistics (1966) which has the significant subtitle A chapter in the history of rationalist thought (see Rosiello 1967). However, since this chapter is dedicated to the (early) history of typology, the problem of mental and linguistic universals and predispositions in the perspective of what could be labeled ‘cognitive psychology’ will be dealt with just inasmuch it has (had) substantial consequences for linguistic typology in its narrow sense. A thorough discussion of this topic would by far exceed the limits of this chapter. What is relevant in the present context is that in the 17th, 18th, and 19th centuries one main problem has been how to cope the essential basic 8 homogeneity of all languages as expression of the human mind with the great differences obtaining among languages. Leibniz (1646-1716), for instance, was looking for a characteristica universalis, a kind of algebraic metalanguage for all humans, where every possible thought could be expressed; the more so as all existing languages derive from a lost original language whose roots are however still present in the languages of the world. At the same time Leibniz was convinced that the ‘differentia linguarum’ is not the consequence of the Babel sin but a necessary consequence of the human nature and its cultural different evolutions. He studied dialects and different languages doing also field research! (see Gensini 1990). The idea of large collections of texts in as many different languages as possible was already present in the 16th cent. The Swiss Konrad Gesner published (1555) his Mithridates, sive de differentiis linguarum tum veterum tum quae hodie (“M., or the differences among the ancient as well as the modern languages”). And Claude Duret wrote between 1613 and 1619 a Thrésor de l’histoire des langues de cet univers. The so-called ‘missionary linguistics’ played a highly important role in making European scholars acquainted with languages of the Americas which had remained unknown until the discovery and conquest of the New World. Also the Jesuits in Asia contributed a lot to the widening of the linguistic horizons. It is impossible to mention all the books which laid the basis of modern comparatism which has been the precondition for the birth of typology, too (cp. on this matter De 9 Mauro / Formigari 1990). But we have to quote at least the famous Catálogo de las lenguas de las naciones conocidas (1800) composed by the Spanish Jesuit Lorenzo Hervás y Panduro (1735-1809) who used the works of his missionary colleagues. 1767 Nicolas Beauzée had published a Grammaire générale ou exposition raisonnée des éléments nécessaires du langage, pour servir de fondement à l’étude des toutes les langues. The title reveals the two aspects of 18th century linguistics. The aim is to arrive at a general, scientific and speculative (raisonnée) theory of language, but to attain this goal it is necessary to know empirically many different languages: and Beauzée mentions in his Grammaire not only Greek, Latin and Hebrew but also Swedish and Lappish, Irish and Welsch, Baskish, Quechua and Chinese (along with Spanish, Italian, German and English). Some years later Johann Christoph Adelung collected a very large language sample which was completed and edited between 1806 and 1817 by Johann Severin Vater after Adelung’s death (1806) with the title Mithridates, oder allgemeine Sprachenkunde mit dem Vater Unser als Sprachprobe in nahe fünfhundert Sprachen und Mundarten (1806-1817). The underlying idea was that the cross-linguistic comparison of very different languages may be able to uncover the general philosophical principles, i.e. the characteristica universalis, and at the same time to recover the evolution of the man’s faculty of language. 10 The allgemeine Sprachenkunde (which could be translated with ‘general linguistics’) had thus the twofold goal, already alluded to: first to describe the many new languages the colonial expansion of the European states became acquainted with, also for practical interests such as christianization or trading; second to arrive at the very nature of the human language, in platonic terms we could say the ‘idea’ of language per se. That’s why ‘linguistic comparison’ meant both philosophical comparison of languages (philosophische Vergleichung) as in Herder’s Ideen zur Philosophie der Geschichte der Menschheit (1784-1785), and comparative grammar as in Vater’s Lehrbuch der allgemeinen Grammatick (1805) (see Morpurgo Davies 1994: 93; Ramat 1990:200). Even in the book which is traditionally considered to mark the beginning of modern linguistics, i.e. Franz Bopp’s famous ‘Conjugation system’ (Über das Conjugationssystem der Sanskritsprache in Vergleichung mit jenem der griechischen, lateinischen, persischen, und germanischen Sprache: 1816), we see that philosophical speculations and historical approach are strictly intertwined (das Sprachstudium als ein historisches und philosophisches zu behandeln, i.e. ‘to treat the study of language both as a historical and philosophical one’, as K.J. Windischmann wrote in his Introduction to the Conjugationssystem). Bopp compared similar forms in order to examine their inner structure, not to reconstruct the original form but to confirm his typological hypothesis that all sentences are formed by a subject 11 and its attribute linked by the verb ‘to be’: homo est mortalis is the prototypical sentence of the Indo-Eruopean languages. Both viewpoints are present in Bopp’s admirer Wilhelm von Humboldt (1767-1835) who can really be considered the bridge among the Enlightenment rationalistic, more philosophical and speculative approach to language(s) and the Romanticism of the first part of the 19th century. The title of his posthumously published most important work, namely Über die Verschiedenheit des menschlichen Sprachbaues und ihren Einfluss auf die geistige Entwicklung des Menschengeschlechtes (1836: ‘On the difference of the human linguistic structure and its influence on the intellectual development of the mankind’), seems to be oriented more toward the relativistic position of the Romanticism which was interested on linguistic diversities as mirror of the spiritual and intellectual differences among cultures. But in many pages of the Verschiedenheit it is clearly stated that all languages are just reproductions (Abbilde) of the human nature with its ability for speaking . We can say, he writes, that the entire human kind has a unique language and that at the same time every human being has his own particular language: “[…] since the natural predisposition towards language exists in general in man and all men must have in them the key to understand all languages, it automatically follows that the form of all languages must be essentially the same […] The difference may only consist in the means and restrictions affecting the possibility of achieving this goal” (remember of Dante’s approach, illustrated in section 2). It 12 is however a matter of fact, that it was precisely these differences and ‘restrictions’ which interested Humboldt most. He was interested in the linguistic structures of the many languages he dealt with, which reflect –he said- their character, or, to use his own word, their genius: under this point of view we may consider Humboldt as a forerunner of ethnolinguistics (on Humboldt’s typology see below).i [space] 5. Meanwhile the new historically oriented comparatism of August Wilhelm Schlegel (1767-1845) and his brother Friedrich (1772-1829), Friedrich Bopp (1791-1867) and Jacob Grimm (1785-1863) had more and more differentiated philosophy and linguistics. Empirical research of purely linguistic facts in the new positivistic attitude won gradually over the philosophy of language. The studies by Friedrich and even more by August Wilhelm Schlegel led to a first typological division of languages. In his Observations sur la Langue et la Littérature Provençales (1818) A.W. Schlegel says that all languages can be divided in three classes: a.) languages without any grammatical structure, like Chinese; b.) languages with agglutinated affixes, like Turkish; c.) languages with inflections, to be distinguished on their turn in c'.) synthetic and c''.) analytic languages. As examples of c'.) Schlegel quotes Latin and Ancient Greek, whereas to c''.) belong languages such as French which, contrary to the classical languages, makes use of articles, personal 13 pronouns before the verb, auxiliaries and prepositions. The Germanic languages are to be located between c'.) and c''.). Schlegel notes that this basic tripartition had been developed by his brother Friedrich ten year earlier (Über die Sprache und Weisheit der Indier [1808], with the significant subtitle Ein Beitrag zur Begründung der Alterthumskunde, i.e. ‘A contribution to the founding of the science of the antiquity’) and was first inspired by Adam Smith’s Considerations concerning the first formation of languages and the different genius of original and compounded languages (1759). Actually, Smith had divided languages in two types: a.) primitive, simple, original, and uncompounded languages ; b.) compounded languages. Ancient Greek is a good example of the first type: its strongly developed inflectional system is self-sufficient in order to express the relations obtaining in the sentence. On the contrary languages such as Italian or French which have a poorer inflectional system must have recourse to phrases (‘compositions’). Basically we have to do with the opposition between synthetic and analytic languages, which will be basic for many further typological researches (cp. Coseriu 1968). From the evolutionary point of view Smith’s idea was that the synthetic type is more ancient than the analytic one –and even preferable. A twofold division had been proposed some years earlier by the abbot Gabriel Girard (1677-1748) in his Vrais principes de la langue françoise (1747). This division is based on syntax rather than on morphology. Girard contrasted the ‘analogous’ languages to the ‘transpositive’ ones. The former 14 include Hebrew, French, Italian and Spanish; they are called ‘analogous’ because their word order is basically SVO, which is analogous, i.e. corresponds to the ‘natural’, logical way of thinking: first comes the Subject then the verbal Predicate and finally the Object: John loves Mary; Bill writes a letter. Languages such as Latin, Old Slavic etc. are on the contrary ‘transpositive’ because they have preferably an SOV order (Brutus Caesarem necavit “Brutus killed Caesar”) which transposes the natural way of thinking (cp. Ramat 1995:29). Finally there are also ‘mixed’ (amphilogique) languages such as Greek and German. Schlegel’s subdivision was basically accepted also by Humboldt, though with an important refinement. He introduced a fourth language type, namely the incorporating (einverleibend) one which unifies in a single word many concepts which in our European languages are necessarily expressed by more words in a sentence. To this type belongs Delaware and other American languages. The introduction of this fourth type represents a shift from a morphologically oriented typology to a more syntactically oriented one (Ramat 1987:206f.). It has to be noted that Humboldt does not classifies languages but refers to language types, i.e. to constructional principles. Anticipating Vladimir Skalička he recognizes that no language belongs to a single type. A ‘linguistic type’ is rather an ideal construction which in no language is ever realized (Coseriu 1973: 253). It is important also to note that in Humboldt’s eyes the types represent different stages of an evolution which, starting by isolated 15 words, arrives finally to the inflectional languages where the previously isolated words are fused (zusammengesetzt) in a more complex form: “according its origin every inflection is the fusion of different signs or, better, words” (Humboldt 1819-20: 158). There is here a clear influence of Bopp’s Conjugationssystem, where the verbal conjugation was explained via the affixation (i.e. agglutination) to the verbal root of the auxiliaries asti and bhavati “to be”. [space] 6. After Wilhelm von Humboldt Indo-European studies became the main trend in linguistics and typology was confined to a more marginal role. The discovery of many ‘scientific’ laws (Lautgesetze), which had already begun with the Danish Rasmus Rask (1787-1832) and Jacob Grimm, became the brand of the so-called Neogrammarians (Junggrammatiker) which dominated the second part of the 19th century. The final point of this evolution may be represented by the famous statement by Antoine Meillet, one of the most prominent heirs of the Neogrammarians in the 20th cent.: historical linguistics is the only useful and valid linguistic classification (Meillet 1914). This statement is contained in an article occasioned by the publication of Franz Nikolaus Finck’s book on the language families of the world (Die Sprachstämme des Erdkreises, 1xxx , which follows Finck 1910 on a very similar matter). In this article Meillet criticizes the canonical classification in 16 four linguistic types, i.e. isolating, agglutinating, incorporating and inflectional languages as useless and non scientific. Long before Finck, August Schleicher (1848) had advanced a tripartite division in monosyllabic, agglutinating and inflectional languages not accepting the fourth type suggested by Humboldt, namely the incorporating one. Later on (1861-62) he refined this tripartition and introduced subclasses by using for every type A. von Schlegel’s synthetic versus analytic criterion. Heymann Steinthal (1860) proposed a new typological classification of the languages, partly diverging from Humboldt’s fourfold division. The subtitle of Steinthal’s book indicates that the Humboldtian philosophical approach (see section 4) to the classification is still operative. Note that Steinthal has been the editor of Humboldt’s linguistic writings (Steinthal 1884). As Humboldt, Steinthal underlines the psychological aspect of language : the evolution of the human mind matches the development of language. At the same time Steinthal is faced with the same problem as Humboldt: why languages are different if the language faculty is a faculty common to the entire human kind? To give an answer to this question Steinthal elaborates again on some humboldtian ideas and develops what will receive the name of ‘ethnopsychology’ (Wundt’s Völkerpsychologie; see Wundt 1912; cf. Graffi 1991:45-50). But already Graziadio Isaia Ascoli had maintained a very sceptical position towards the so-called ‘psychological grammar’ based on “speculative observations of the 17 different linguistic types in order to provide a firm basis to a psychology of nations” (Ascoli [1866 >]1877: 42-44). As for the languages’ classification, Steinthal states that it is necessary to find feature(s) which may characterize and determine the entire ‘organism’ of a language (on holistic typology see section 7, below). He introduces the concept of ‘form’ and distinguishes two main types: a.) languages without form (formlose Sprachen), i.e. without formal expression of the grammatical relations: these are expressed by full words: e.g. the plural is expressed by words meaning “many, all”, tenses are expressed by particles such as “time ago” etc. Siamese and Birman are representatives of this type. Chinese, contrary to the received classification, belongs to the languages endowed with form (Form-Sprachen: type b.)) since grammatical relations are expressed by the words juxtaposition (Nebensetzung) in the sentence. The best examples of the Form-Sprachen are however the Semitic and, above all, the Indo-European languages which modify (abwandeln) the form of the word. Note in any case that juxtaposition and modification can be found in both types. Thus, Turkish and, more generally, the Uralo-Altaic languages are ascribed to the a.) type, but they have derivative suffixes like the Indo-European languages which belong to type b.). Similarly, Siamese and Birman (type a.)) make use of the Nebensetzung and so does Chinese (type b.)). Steinthal’s classification, based on ‘ethnopsychology’, is not without contradictions (see Ascoli’s criticism on the very diffused ‘gazzarra psicologica’ (psychological din) of his time). 18 Franz Misteli revised Steinthal’s typology (Misteli 1893) and modified the opposition ‘formless’ vs. ‘form languages’ into ‘non-word’ vs. ‘word languages’. The classifying parameters are the relation of the word to the sentence (syntactic criterion) and the inner structure of the word (morphological criterion). Misteli’s concept of ‘word’ is very strict and echoes Humboldt’s ideas. According to Humboldt, only the inflectional languages fully realize in the word the inner fusion of meaning and syntactic function: the synthesis realized in the word unity is the most brilliant solution, since by using this strategy every element of the sentence shows at the same time its objective meaning and its relation to the underlying thought. Thus, the ‘best’ word has to have some internal mutation such as the Ablaut or the Semitic vowel insertion in the three consonantal root. On the contrary, agglutinating languages simply add more elements (suffixes) to an unchangeable root in a mechanic way. Using these parameters and building on Steinthal’s typology Misteli proposed the following subdivision: a.) Formless languages i. with sentence-words (Ein-Wort-Sätze, i.e. the incorporating type of the Amerindian languages); ii. without words (nichtwortig) α. root-isolating such as Chinese β. stem-isolating such as Malay 19 γ. juxtaposing (anreihend) such as Egyptian; iii. with apparent words (scheinwortig) such as Turkish b.) Form languages with real words (echtwortig): Semitic and Indoeuropean languages. We have already alluded to Finck (1867-1910), whose typology close parallels Misteli’s classification. His definition of word is the following: “the smallest constituent of the utterance, which is not bound in a rigid manner to other phonetic clusters”. On the basis of this definition he distinguishes between languages having a high complexity of the word such as the Inuktikut and the other polysynthetic languages, and languages such as the Bantus whose words are ‘fragmentary’ structures and consist of only loosely related parts. The other languages of the world are located along a continuum between these two poles. Chinese, where each word is a single morpheme, is deemed to be a conceptually simple language, representing more or less the center of the continuum. Finck’s three main groups are: a.) isolating ; b.) inflectional; c.) combining, but not inflected languages (such as Turkish or Inuktikut). We disregard here the subdivisions of a.) – c.) suggested by Finck, which are not very different from those operated by Misteli. Misteli’s and Finck’s suggestions were not accepted by the largest majority of linguists (see the previously quoted sharp refusal by Meillet), also due to the attempts of the ‘ethnopsychology’ to find a link between race and 20 language. The traditional fourfold division based on Schlegel’s and Humboldt’s suggestion became the typology ‘vulgata’. Finally we have to mention also Max Müller (1823-1900) who returned to Bopp’s approach that considered roots as the basis for typology. Isolating, agglutinating or inflectional languages represent but different degrees of a unique strategy of composition between predicative and demonstrative roots. The fourth traditional type, i.e. the polysynthetic, is not necessary, since it represents just a particular case of composition, and not a different grammatical strategy (see Müller 1880). [space] 7. Georg von der Gabelentz (18xx-19xx) can be considered as the final moment in our sketch of the early history of typology. Not only because he probably used first the word ‘typology’ in his well-known and often quoted page from the 2nd edition of his book Die Sprachwissenschaft (Gabelentz 1901:481)ii : “but what an achievement would it be were we able to confront a language and say to it: ‘you have such and such a specific property and hence also such and such further properties and such and such an overall characther’ ─ were we able, as daring botanists have indeed tried, to construct the entire lime tree from its leaf” (transl. by Shibatani and Bynon 1995: 10). And the text continues: “If one had to baptize a not yet born child, I would choose the name t y p o l o g y. I see here a task for general linguistics, whose solution can already be tempted with the means we have now at our disposal”. As Jae Jung 21 Song (2001:358) correctly suggests, the sentence by Gabelentz , taking as scientific model postdarwin botany, envisages a holistic typology on the basis of a unique overarching principle or, to use an expression of W. P. Lehmann, of very few underlying ground-planes. A special issue of a linguistic journal was in fact devoted some years ago to comment Gabelentz’ wishful perspective (Folia Linguistica 20, 1 – 2 (1986), guest-editor Frans Plank). The discussion of this important topic belongs to the following chapters; but one must not forget the remarks with which Louis Hjelmslev concluded the chapter on the typological relations between languages of his Sproget (1963): “An exhaustive linguistic typology is, in fact, the biggest and most important task facing linguistics [….] Its ultimate aim must be to show which linguistic structures are possible, in general, and why it is just those structures, and not others, that are possible. And here it will come closer than any other kind of linguistics to what might be called the problem of the essence of language. [….] Only through typology does linguistics raise to quite general points of view and become a science” (Hjelmslev, Engl. transl. : Language, 1970 : 96). It seems to me that the same need for a general overview and a possible all-embracing explanation constitutes an important link uniting Gabelentz to Hjelmslev. And also some other points deserve to be underlined in Gabelentz’ typological views, that show his very modern approach ─ so that it is correct to consider Gabelentz as the bridge between the 19th and 20th century typologists. 22 He clearly envisaged that , if typology has to achieve reliable, non impressionistic results, it has to make use of questionnaires (see above section 4. on Leibniz) prepared by experts, with a fine-grained research program “so that any question may be answered by ‘Yes’ or ‘No’ “ (which is not always the case….: P.Rt.). Statistics should then be used in analysing the data: this would finally enable researchers to overcome the many contradictory assessments which can be found in the literature concerning the presence of this or that linguistic feature. The presence of features α and β in a given language will entail with more than chance probability ─to use Greenberg’s famous expression ─ the presence of features γ and δ. It can easily be seen that Gabelentz’ program is really very modern: typology aims at becoming predictive. Another very modern point in Gabelentz’ view is that, contrary to the largest part of his predecessors, he considers linguistic change and linguistic type change in a really unprejudiced way. He does not think of diachronic evolution as language decay (Sprachzerstörung according to A. Schleicher), nor as continuous improvement (a position held, e.g., by Otto Jespersen) but as a spiral movement (Spirallauf : Gabelentz 1901: 255-258) that may come back to previous typological stages ─though using different linguistic means. Agglutinated suffixes undergo phonetic erosion and disappear; their functions are taken over by the syntactic word order and full words (: isolating languages). But on their turn the full words (remember the examples of Steinthal’s formlose Sprachen in section 6) may be agglutinated 23 and, again, be phonetically eroded and disappear. Consequently, new periphrastic forms are preferred, which may again be contracted in a single word (e.g., Lat. *vidē-fuō > vidēbō, late Lat. videre habeo > It. vedrò, Fr. je verrai, Span. veré, whereas Engl. I shall see is a periphrastic form). As can be seen from this example, Gabelentz does not propose new typological classifications: his view is the canonical one, with two poles represented by the isolation and the incorporating polysynthesis, respectively. His relevance lies elsewhere: it is a matter of fact that the idea of a spiral movement in language change has been taken up by many contemporary linguists as a very handy representation of the moving forces which cause a language or a linguistic type to drift from stage A to stage B and, eventually, to stage A'. [space] 8. A final comment can be done as a summary of this chapter. It is true that today’s typology has adopted induction as its general strategy: starting from the comparison of large and representative language samples, chosen by means of statistical criteria, it aims at establishing the variation range among languages (and consequently what are the universals of language: remember Hjelmslev’s word). Latin is no longer the model which has deductively to be applied to the description of other languages. Nevertheless the history of typological studies does not show the sharp contraposition between nineteenth century typology based on morphology, not capable of illuminating the inner structure of 24 language and , on the other hand, twentieth century typology based on syntax (from Greenberg on), as maintained for instance by Winfred P. Lehmann (1993). Speaking of Humboldt, Bopp, Steinthal etc. ─and even earlier of the Modistae─ we have seen that grammatical considerations were strictly intertwined with reflection on syntax. We can speak of a morphosyntactic approach to typology. Phonology and semantics, on the contrary, have not been paid the same attention in the period considered here (nor, to tell the truth, in the next periods). 25 Key words:, Comparatism, Enlightenment, Ethnopsychology, Holistic typology, Language classifications: -Agglutinating -Analogous -Analytic -Compounded -Formless -Incorporating -Inflectional -Inorganic -Organic -Synthetic -Transpositive -Uncompounded -With form Modistae, ‘parts of speech’, Philosophy, Positivism, Renaissance, Romanticism, Scholastica, Universals of language Items for further reading: • Comrie, B. (1989). Language universals and linguistic typology. 2nd ed. Oxford: Blackwell. • ‘De la grammaire à la linguistique’. Histoire, épistémologie, langage. Tome 3, fasc. 2 , 1981. Presses Universitaires de Lille. • De Mauro, T., and Formigari, L., eds. (1990). Leibniz, Humboldt, and the origins of comparativism. Amsterdam / Philadelphia: Benjamins. • Droixe, D., coordin. ‘Genèse du comparatisme indo-européen’. Histoire, épistémologie, langage. Tome 6, fasc. 2 , 1984. Presses Universitaires de Lille. • Greenberg, J.H. (1974). Language typology: A historical and anlytic overview. The Hague: Mouton. 26 • Haspelmath, M. , König, E., Oesterreicher, W., and Raible W. (eds.) (2001), Language typology and language universals. Berlin and New York: de Gruyter, vol. I, 1.iii: ‘History and prehistory of universals research’; iv: ‘History and approaches of language typology’. • Jung Song, J. (2001). Linguistic typology. Morphology and Syntax. Harlow etc.: Longman. • Lepschy G.C., a c. di, (1990-1994) Storia della linguistica, 3 volumes. Bologna: Il Mulino [Engl. transl.: History of linguistics. Harlow etc.: Longman 1998]. • Massariello Merzagora, G., a c. di, (2000). Storia del pensiero linguistico: linearità, fratture e circolarità. Atti del convegno della Società di Linguist. Ital. (Verona 1999). Roma: Il calamo. • Robins, R.H. (1997). A short history of lingustics. 4th ed. London: Longman. • Rosiello, L. (1967). Linguistica illuminista. Bologna: Il Mulino. ● Shibatani, M. and Bynon, Th., eds. (1995). Approaches to language typology. Oxford: Clarendon Press. References: *Ascoli, G.I. (1867). Review of G. Lignana, La grammatica comparata di Bopp (Napoli 1866) [Repr. in G. I. A., Studj critici II, Roma etc.: Loescher, 1877: 31-61] *De Mauro, T., and Formigari, L., eds. (1990). Leibniz, Humboldt, and the origins of comparativism. Amsterdam / Philadelphia: Benjamins. 27 *Coseriu, E. (1968). ‘Adam Smith und die Anfänge der Sprachtypologie’, in H.E. Brekle and L. Lipka (Hsgb.), Wortbildung, Syntax und Morphologie. Festschrift Hans Marchand. The Hague: Mouton, 46-54. * ----------- (1973). ‘Sulla tipologia linguistica di Wilhelm von Humboldt. Contributo alla critica della tradizione linguistica’. Lingua e Stile 8: 235-266. * Finck, F.N. (1910) Haupttypen des Sprachbaus. Leipzig: Teubner. * Gabelentz, G. von der (1901). Die Sprachwisenschaft. Ihre Aufgaben, Methoden und bisherige Ergebnisse. Leipzig: Tauchnitz (2., vermehrte und verbesserte Auflage, hsgb. von A. Schulenburg [Repr. Tübingen: Narr 1972]. * Gensini, S. (1990). ‘ 'Vulgaris opinio babelica'. Sui fondamenti storico-teorici della pluralità delle lingue nel pensiero di Leibniz’ in De Mauro, and Formigari : 61-83. * Graffi, G. (1991). La sintassi fra Ottocento e Novecento. Bologna: Il Mulino. [Engl. edit.: 200 years of syntax. A critical survey. Amsterdam and New York: Benjamins]. * Hjelmslev L. (1970) Language [Dan. Original : Sproget. En introduktion. Engl. transl. by Fr. J. Whitfield]. Madison : University of Wisconsin Press. * Humboldt, W. von (1819-20). Gramática vasca. [Unpubl. German text, transl. into Span. by J.Garate, Guillermo de Humboldt. Estudios de sus trabajos sobre Vasconia. Bilbao 1933]. *Jung Song, J. (2001). Linguistic typology. Morphology and Syntax. Harlow etc.: Longman. *Lehmann, W.P. (1993). ‘Diagnostic uses of typology’, Languages of the world 7,2: 3-13. 28 *Matthews, P. (1990). ‘La linguistica greco-latina’, in G.C. Lepschy , Vol. I: 187310. *Meillet, A. (1914). ‘Le problème de la parenté des langues’. Scientia 15 [repr. in A. M., Linguistique historique et linguistique générale. Paris: Champion, 1975: 76-101]. * Misteli, Fr. (1893). Charakteristik der hauptsächlichsten Typen des Sprachbaues. Neubearbeitung des Werkes von Prof. H. Steinthal. Berlin: Dümmler. * Morpurgo Davies, A. (1994). ‘La linguistica dell’Ottocento’, in G.C. Lepschy , Vol. III: 11- 399. * Müller, M.(1880). Lectures on the science of language. Vol. I. 6th ed. London: Longmans. *Plank, Fr. (1991). ‘Hypology, typology: The Gabelentz puzzle’, Folia linguistica, 25: 421-457. *Ramat, P. (1987). ‘The language typology of Wilhelm von Humboldt’, in P. Ramat, Linguistic typology. Berlin / New York / Amsterdam: Mouton de Gruyter, Chapt. 10. *------------ (1990). ‘Da Humboldt ai neogrammatici. Continuità e fratture’, in De Mauro and Formigari : 199-210. *------------(1995). ‘Typological comparison: Towards a historical perspective’, in Shibatani, M. and Bynon, Th. : 27-48. *Rosiello, L. (1967). Linguistica illuminista. Bologna: Il Mulino. * Shibatani, M. and Bynon, Th., eds. (1995). Approaches to language typology. Oxford: Clarendon Press. * Schleicher, A. (1848). Zur vergleichenden Sprachengeschichte. Bonn: König. 29 *-------------- (1861-62). Compendium der vergleichenden Grammatik der indogermanischen Sprachen. Kurzer Abriss einer Laut- und Formenlehre der indogermanischen Ursprache, des Altindischen, Altiranischen, Altgriechischen, Altitalischen, Altkeltischen, Altslawischen, Litauischen und Altdeutschen. 2 vols. Weimar: Böhlau [Engl. transl. by H. Rendall. London: Trübner 1874-1877]. *Steinthal, H. (1860). Charakteristik der hauptsächlichsten Typen des Sprachbaues. Zweite Bearbeitung seiner Classifikation der Sprachen (dargestellt als die Entwicklung der Sprachidee ) [1850]. Berlin: Dümmler. * ------------- (1884). Die sprachphilosophischen Werke Wilhelms von Humboldt, herausgegeben und erklärt von Dr. H. Steinthal. Berlin: Dümmler. * Tavoni, M. (1990). ‘La linguistica rinascimentale’, in G.C. Lepschy ( a c. di), Vol. II: 169- 312. *Vineis, E., and Maierù A. (1990). ‘La linguistica medioevale’, in G.C. Lepschy , Vol. II: 11- 168. * Wundt, W. (1912). Völkerpsychologie. I. Die Sprache. Leipzig: Engelmann. #################################################################### Note for Jae Jung Song: This file has the following word total number (including ‘items for further reading’ and bibliography) :6.650 30 i The notion of ‘genius of language’ would deserve a detailed discussion, which can just be summarised here (see Rosiello 1967 : 79-87; W. Hüllen in Haspelmath et al. 2001: 242 f.). The term, probably introduced by Amable de Bourcey in a discours before the French Academy (1635), had a period of fashionable usage in the 17th cent. which shows an almost ridiculous arbitrariness of judgement. Spanish is considered a “langue orgouilleuse”, Italian “une langue coquette” and French is said to be ‘prude’. Though totally deprived of any non-impressionistic criterion, such judgements, which unfortunately still form part of the prejudices through which many Europeans look to each other, are a first step to link anthropological, cultural, historical, and linguistics facts along the path ethnolinguists will later try in a much more serious manner, in the wake of Humboldt up to Sapir and Whorf. This appears evident already in Condillac’s Essai sur l’origine des connoissances humaines (1746): by ‘génie de la langue’ is meant the particular system of semiotic signs the language of a particular nation makes use of . Language is thus the picture of the character and genius of the nation speaking it. ii On the philological question whether the term really goes back to Georg von der Gabelentz or to the editor of the 2nd edition (1901), namely his nephew Albrecht Conon von der Shulenburg, see Plank 1991.