„Tuščia vieta“ yra metodologiškai pažeidžiama sąvoka

162 | PASTABOS More varied fluctuations are found in the Akmenė, Klykoliai, Papilė, Vegėriai, and Tirkšliai dialects (the abbreviation Vg (pp. 92–93) used to mean Vaiguva, but this phenomenon does not occur there—there must be some error). For the most part, such stresses are everywhere non-retracting when they occur in modifiers, and sometimes when they occur in a short ending (but pùo žemaičiùs Rdn, Trš, Šiaudinė near Pp; pùo agl’ik’ès Krš, etc.). Šiek tiek pavyzdžių pateikta naujojoje Lietuvių kalbos tarmių chrestomatijoje (2004): Darbėnuose ẹ(š) šèrdė ~ iš širdie (transponuota iš širdie); ẹš jègâ· (transponuota iš jègâi); ar ne ẹš jie-gâ· (~ iš jègai) – negi iš tvirtapradės galūnės galėtų kirtį šitaip atitraukti! Tai nelaukta, nes susiję su iškilia mokslininko kalbininko, konsultavusio šį leidinį, asmenybe. Vytautas Vitkauskas Lietuvių kalbos institutas P. Vileišio g. 5, 10308 Vilnius, Lietuva Received 2005 05 23 “EMPTY PLACE” IS A METHODOLOGICALLY VULNERABLE CONCEPT1 In his response, Can an empty place mean anything? (Acta Linguistica Litu-hanica 50, 136–138), Vytautas Ambrazas, discussing my views on the so-called zero copula, maintains that my “logical statements are correct, but they do not at all contradict the conception of zero presented in Grammar” (p. 136), since there “zero is regarded as a methodological concept that helps determine when the absence of a certain marker has a defined grammatical function” (p. 137). If zero (or an empty place) belongs to methodology, then it should be used as is customary in other sciences—“unburdened” by semantics. I myself do not object to using zero as a purely technical device (cf. the explanation of vowel gradation, which uses the term zero grade). Nor do I object to the term empty place, since, for example, formal grammar theory uses an empty (basic-element) string, whose symbol E is rendered as ‘ ’. I object only to assigning meaning to an empty place, because such grammars are set apart from all empirical sciences (I cannot regard grammatical facts as merely components of the psyche, since linguistics is not a branch of psychology for me). In presenting my reasoning about the so-called zero copula, I have already shown how a logical contradiction arises: a meaning that is not communicated by the text is ignored, and an additional meaning is attributed to the text (through overinterpretation). I think this often happens because 1 When I found the article Lietuvių kalbos enciklopedijoje, published in 1999, entitled Nulinis kalbos elementas (p. 441), and, after giving it careful thought, realized that I could not agree with most of its claims, I decided to set out my reasoning within a short time in three texts: (i) on the so-called zero copula, (ii) on so-called zero endings, and (iii) on so-called nominative sentences (such as, for example, Žiema.), to which some also attribute zero expression of the verb būti. Only then would my view of “zero” elements of language become clear in full, and only then did I expect to take part in a discussion, if anyone disagreed with my reasoning. But after I published my thoughts on the so-called zero copula (Acta Linguistica Lithuanica 48, 177–184), various misfortunes completely interrupted my reflections, and in the meantime I received a polemical response from Vytautas Ambrazas. Consequently, temporarily setting aside my earlier plan, I am forced to enter the discussion. I tell readers of this text that I am reserving some arguments concerning zero elements of language for later, since I hope to write about both zero endings and so-called nominative sentences. 2 Гросс, М., Лантен, А., Теория формальных грамматик, Москва, 1971, 15. PASTABOS 163 principles of semantic analysis that are poorly understood. Therefore, I will discuss once again the example from the Grammar The weather is nice / The weather is nice (p. 612). These two sentences are considered absolute synonyms, since otherwise one would not say that “such a zero [...] copula is equivalent to the present-tense form of the verb to be” (ibid.). The sentence The weather is nice (as a speaker of the Utena dialect, I use sentences like this myself very often) describes a state of the atmosphere that coincides with the time of utterance. This can also be established through objective observation from the outside, rather than by analyzing mental content. More than one linguist lacking a clear understanding of semantic analysis has already concluded from this that the sentence The weather is nice has present (present-tense) meaning. This is precisely where one must ask: where does such information come from? In this case, the observed fact (the present) belongs to the conditions of use of the sentence, but it does not follow that this information is also expressed in the text, since it may be unexpressed4. This is so for at least three reasons: (i) language is based not on the laws of nature, which are necessary and universal, but on social habits (it is conventional); therefore, not only may something we consider important be absent from different languages5, but it may also vary among the dialects of one and the same language; (ii) it should not be forgotten that texts are not exhaustive representations of reality, because even long texts do not cover everything; they are the result of selection; (iii) and a speaker may see no need to say something (in the case under discussion, one might suppose that the speaker leaves out “is” because it is obvious, but then sentences with is would have to be treated as basic and one would have to speak of ellipsis, an account I cannot support (this will become clear shortly). It is clear that neither weather nor nice has tense meaning, and the so-called “empty slot,” being merely another name for the empty set, which contains nothing (no element)6, cannot have tense meaning either, since in that case nothing would already turn into something7. Syntacticians usually examine selected sentences and therefore cannot see them in a broader context, so I will provide sentences of the type under discussion with (minimal) context: Pine trees grew green all around. The forest is calm, quiet.Only here and there an old woodpecker taps. Morning dawned, the day grew bright. / I was so exhausted that I began raving at night. The weather is cold, especially at night. / 3 Following Leon Zawadowski, who greatly influenced my thinking, I call the constitution the speech situation, and I consider this term necessary (see Zawadowski, L., Lingwistyczna teoria języka, Warszawa: Państwowe Wydawnictwo Naukowe, 1966, 272–273). 4 We do not infer from reality what must be expressed in a text; rather, (normally) we draw conclusions about reality from the text. I think I explained this well when presenting Leon Zawadowski’s views on language (see Kalbotyra 24 (1), 1972, 111–119), and I do not consider myself responsible for other linguists not having taken an interest in it. 5 For example, Indonesian undoubtedly has verbs, but they have neither tense nor mood categories, so the sentence Anak tidur can mean The child is sleeping, The child slept, or The child will sleep, and it would be a serious mistake to assign the meaning of the verb tidur on the basis of its relation to the time of utterance. 6 On the concept of a (mathematical) set, see Visuotinė lietuvių enciklopedija 1, Vilnius: Mokslo ir enciklopedijų leidybos institutas, 2001, 180 (entry “aibė”). It would suffice for linguists to consult: Bulota, A., Susipažinkime su aibėmis, Vilnius: Mintis, 1973 (especially section 11). Naumas Vilenkin explains well why mathematics needs the empty set (Виленкин, Н. Я., Рассказы о множествах, Москва: Наука, 1965, 27–28). 7 The thinking of linguists who posit “zeros” strongly reminds me of the anti-logic and anti-scientism of Martin Heidegger, whose notorious saying Das Nichts nichtet (roughly: “Nothing manifests itself”) became a target of criticism by the Vienna positivists (and of mockery of Heidegger’s manner of speaking more generally). For Heidegger’s language, see Carnap, R., Metafizikos įveikimas kalbos loginės analizės būdu. In: Nekrašas, E., Filosofijos įvadas, Vilnius: Mokslo ir enciklopedijų leidybos institutas, 2004, 277–281. (I myself, of course, side with Carnap, not Heidegger.) 164 | P A S T A B O S I got on a suburban train. The carriage was long. [...] It was night, completely dark. | We arrived in Dolgomostas. The road was very bad. There were already Lithuanians brought here earlier. These excerpts were selected from the memoirs of former prisoner V. Ulčinskas8, written forty years later, so there can be no question here of the coincidence of the verbless sentences with the time of writing. As is customary, memoirs speak of the past in the past tense. If the isolated verbless sentences in this case, too, have present-tense meaning, then we would have to acknowledge that the narrator, through inexperience, keeps shifting the tense of the narrative. I do not think so, because V. Ulčinskas is a good storyteller. If in these cases the so-called zero copula has the meaning of the sentences in the context, then such sentences are equivalent to sentences with was, which means that they have another grammatical meaning in addition to the present-tense one. I do not support this claim either: I think that in this case the descriptive verbless sentences are inserted into narrative sentences depicting the past. One can give another argument that the sentences under discussion do not have present-time meaning. Here is a grammatically perfectly correct sentence: The weather in Salton Sink (USA) is always beautiful. How would proponents of present-time meaning explain, in the sentence The weather is always beautiful, the use in this case of the adverb always, which expresses timelessness? Whether the sentence I used, My brother is disabled, can be explained as “a member of the paradigm of sentences containing a copula Brother is / was / used to be / will be disabled, used with present-tense meaning” (p. 137) depends on how the paradigm is understood. Linguists use the concept of paradigm without a more precise definition for now9, although it should be clear that it cannot be handed down from heaven. I cannot include the sentences Brother is disabled and Brother disabled in the same paradigm, because these two sentences do not form an opposition (and linguists use the very important concept of opposition rather imprecisely). To make clear why I proceed in this way, I will take examples from phonology, because in all the languages with which I am somewhat familiar, phonological systems are the simplest and easiest to analyze. For example, in Lithuanian bado and pado form an opposition, and this opposition makes it possible to distinguish the contrast /b/: /p/ and establish that these two sounds differ as voiced and voiceless. I now ask whether bado (and pado) form an opposition with ado? If we answer yes, we will get the contrasts /b/: /ø/, /p/: /ø/ (taking other pairs, e.g., sako and ako, tako and ako, we will get many more oppositions of the same kind). Thinking logically, we will inevitably have to include a zero phoneme in the inventory of phonemes. There would be no benefit in comparing all phonemes with zero, because real phonemes would differ from the zero one in every acoustic feature, whereas what matters to the phonologist are only the differentials. As far as I know, most phonologists object to the concept of a zero phoneme, although I have not found in their objections the kind of justification I have presented here. Agreeing with André Martinet, I think that the concept of contrast applies to syntagmatic relations, and that of opposition to paradigmatic ones10. Since the relation bado: ado, in my view- 8 Ešelonų broliai, Vilnius: Vyturys, 1991, 183–241. 9 As the Russian mathematician and philosopher of science Georgij Ruzavin rightly observed, in many theories in the humanities, especially at their initial stage of development, fundamental principles are not explicitly distinguished, although they are presupposed (Рузавин, Г. И., Научная теория: Логико-методологический анализ, Москва: Мысль, 1978, 15). Ever since I read the book Логика научного исследования, Москва: Наука, 1965, I have long been convinced that every system of thought worthy of the name of a true theory, however much some may dislike it, must constantly be reviewed and refined. 10 I take this thought of André Martinet from: Girdenis, A., Teoriniai lietuvių fonologijos pagrindai, Vilnius: Mokslo ir enciklopedijų leidybos institutas, 2001, 66 (footnote 24). PASTABOS 165 mu, belongs to contrast, not opposition; therefore, given that the most important concepts must be applied uniformly to all levels of language, the sentence Brolis neįgalus ‘Brother [is] disabled’ cannot be included in a paradigm with the verb būti ‘to be’. I would point out that on this question I fully agree with Émile Benveniste, who stated very clearly in his widely known article “La phrase nominale” that the study of the nominative sentence must be entirely separated from the study of sentences with būti ‘to be’, that no zero can be discerned in it, and disagreed with Louis Hjelmslev, who identified three implicit elements in the Latin sentence Omnia praeclara rara: the infectum, the present tense, and the indicative mood11. Vytautas Ambrazas also attempts to support his position—the claim that an empty position has meaning—with diachronic arguments. As a proponent of functional linguistics (though I taught Lithuanian historical grammar for many years), I am firmly convinced that when describing language synchronically as a functioning system, diachrony and its attested changes must be left aside, because a synchronic description of language is epistemically primary, whereas a diachronic one is secondary: after all, one cannot observe changes without first determining what is changing12. I cannot agree with Vytautas Ambrazas’s idea of the resemanticization (reanalyzing, see p. 137) of verbless sentences, because Hittite Attaś assuś13 and Lithuanian Tėvas geras, vedu Tvám Viṣṇuh14 and Lithuanian Tu Višnus, pali Dhammā aniccā15 and Lithuanian Dhamai nepastovūs have the same structure and grammatical meaning, and therefore must be interpreted alike: they are descriptive sentences, not sentences expressing “eternal truths” (when writing about the so-called zero copula, I had already noted that sentences with yra ‘is’ may compete with copula-less sentences, but this is not a case of reanalysis). I think we have preserved the old Indo-European model quite well, and that only linguists who improperly operate with zero, going against common sense, become confused in their own thinking and may mislead others. The second point is this: diachronic processes cannot be described in terms of laws, provided we agree that by a law we mean something necessary and universal16. But where there are no laws, there are no changes explainable by laws either. It would be very worthwhile for linguists, as well as all other scholars in the social sciences, to reflect carefully on Karl Popper’s idea that we can observe directions or tendencies of change, but “laws and directions are essentially different things”17; therefore, necessary conclusions cannot be drawn from observing directions or tendencies. And Vytautas Ambrazas, when discussing such sentences in other Indo-European languages18, seems to want to place Lithuanian among the other languages too, but surely… 11 I can rely only on the Russian translation: Бенвенист, Э., Общая лингвистика, Moscow: Progress, 1974, 167–183. 12 See Zawadowski, L., op. cit., 47 and Girdenis, A., op. cit., 28–30 (and footnote 10). 13 For example, from: Иванов, В. В., Хеттский язык, Moscow: Izdatel’stvo vostochnoy literatury, 1963, 171. 14 Pavyzdys iš: Елизаренкова, Т. Я., Ведийский язык, Москва: Наука, 1987, 158. 15 The example is taken from: Елизаренкова, Т. Я., Топоров, В. Н., Язык пали, Moscow: Nauka, 1965, 124. 16 As long as Marxism (now relegated to the sidelines) kept hammering into everyone’s heads that there were even universal laws of human development, it was convenient for linguists, too, to speak without due reflection of laws governing the development (or functioning) of language. 17 Popper, K. R., Rinktinė, Vilnius: Pradai, 2001, 395. 18 Anyone interested in similar Russian sentences would do well to familiarize themselves with the concept of non-localization in time and with the terms “wartość miejscowa w czasie” and “Zeitstellenwert” coined by the German Slavist Erwin Koschmieder, since these ideas may also prove useful in examining temporal meaning in Lithuanian (see Теория функциональной грамматики: Введение. Аспектуальность. Временная локализованность. Таксис, Leningrad: Nauka, 1987, 210–226, 230–232). 166 | P A S T A B O S The use of sentences without a conjunction differs considerably even among Lithuanian dialects, which are genuinely alive and are the only ones to continue the past19. In concluding my discussion of sentences without a conjunction, I want once again to emphasize that the concept of an empty slot (with a positive meaning assigned to it!) arose from an insufficient and superficial analysis of semantics, which mistook the conditions of language use for the meaning of a text. I find it rather hard to understand why Vytautas Ambrazas, a well-known linguist, at the end of his polemic with me acted in a manner quite unlike that of a serious scholar: he began “firing back” at individual objections I had raised to psychologism, although, focused as I was on examining the “zero copula,” I had stated as clearly as possible that my view of psychologism was drawn primarily from Leon Zawadowski, and had indicated where one could learn much more about it. If Vytautas Ambrazas wants to argue about psychologism in earnest, he should argue with someone other than me20. Since, for some reason, Vytautas Ambrazas did not wish to consult the primary source, in some instances he simply drew mistaken conclusions; for example, antipsychologists do not deny that the content of language in the normalu case corresponds to facts of the psyche, but what matters most to them, as functionalists, is not the content of the psyche, which is always individual and cannot pass into the external world—extra corpore—but the meaning of the text, which, being conditioned by social usage, may differ greatly even from the understanding of individual persons. There is another reason: listeners often fail to distinguish meaning from sense (“What did he mean by saying that?”), but in my opinion linguists must make this distinction21. Vladas Žulys 1–7 Tako Street LT-53356 Akademija, Kaunas District, Lithuania Received 14 September 2005 19 Strictly speaking, the standard language cannot simply be fitted into a diachronic process, because it is more or less artificial and may contain all kinds of variation resulting from the influence of dialects. For this very reason, the use of the most frequent Lithuanian verb, būti (‘to be’), should first be investigated in the dialects, and only then should we consider how it is used in the standard language. I do not think its use can be standardized at present. 20 Having graduated in 1959 from the provincial Vilnius University, where no course in logic whatsoever was taught, I can hardly compare my preparation with that of Leon Zawadowski, who studied not only linguistics but also epistemology in the West (I greatly value this fact, since, as I sometimes discover, most other prominent linguists of the twentieth century lack the necessary preparation in epistemology, that is, the methodology of science). Moreover, I think my intellect is weaker than Leon Zawadowski’s, because it has not been well trained to think logically and clearly (it should not be forgotten that linguists in Lithuania were trained only as philologists). 21 Anyone wishing to rid themselves of psychologism, which often does not clearly come to the surface, should consider three cases of language functioning. (i) A message exists, but nobody perceives it. Such a situation may arise when, for example, news is broadcast on the radio but there is no one in the room, or when writings remain unseen by anyone for more than a thousand years (such was the fate of the Bogazköy archive with its Hittite cuneiform tablets). In this case, texts have a representational function, since they always speak about something (one of the principles formulated in the information sciences that emerged in the second half of the twentieth century states that information is always transmitted through a material medium). But in this case language does not perform a communicative function, which rests on the representational one. (ii) A message is recognized as language, but the language itself remains incomprehensible. This happens when people do not know one another’s language or cannot read a particular script (for example, the Phaistos Disc). In this case, too, texts have a representational function, but it is impossible to make use of the communicative function because of lack of knowledge. (iii) Communication takes place normally through language, whether spoken or written, and such language performs both of its principal functions.