Göran Hammarström

Reply to Johannes Kabatek’s comment on Göran Hammarström’s contribution


Kabatek discusses some critical remarks I had made on his invitation. I believe that the points are important and worth being discussed further. Philosophers and linguists can certainly use final and finality with another meaning than it has in common language. However, if one used intentional and intentionality instead, one would avoid an ambiguity that could result in problems for beginners in linguistics and other interested people. Kabatek mentions the possibility of linguistics adopting standards accepted in sciences. Linguists can, however, only use the best standards accepted in language studies. I therefore said that I did not believe in adopting the standards of sciences. Kabatek says now that neither did he believe that these standards should be adopted. He says that he had heard from colleagues “working with quantitative and psycholinguistic methods” that there had been such progress in their studies that we can now be more objective and less intuitive. This is nonsense on two points. (1) Quantitative, or statistical, methods have very little or nothing to contribute to the problems usually discussed in linguistics. However, frequency, transitional probabilitiy, redundancy, predictability and amount of information can be interestingly used in syntagmatic studies but this is almost unknown to linguists. I have used these notions to suggest what I in later publications call prompts (not causes) for the disappearance of phonemes between Latin and French (see Sound change and information theory in H. Eichner et al., eds, Fremd und Eigen [...] in memoriam Hartmut Katz. Wien: Edition Praesens 2001, p. 33–37). I have also used these viewpoints to elucidate the difference between uni and university in my comment on Willems’ contribution. Psycholinguistic methods can, however, add some additional understanding of language. (2) The most important way of understanding language has been and remains intuition. Itkonen has explained that linguistics is basically a hermeneutic science. I have tried to define empirical adequacy in linguistics in several publications (for instance in Fundamentals of Synchronic Linguistics. München: Lincom Europa 2012, p. 4–5) and an important point is that I believe that intuition and introspection have a central place if one wants to investigate language in an adequate way. The whole content side of language is unavailable to any “objective” study.

The discussion about Saussure’s dichotomy synchrony – diachrony never stops because any number of linguists, including Coseriu and Kabatek, refuse to understand the problem correctly. Saussure is right. Firstly, one must understand that when Saussure separates the two, it cannot mean that diachronic study is not based on the comparison of two or more synchronies. There is thus synchrony in diachrony. That is known by Saussure and all other linguists. It is not a problem. Secondly, the crucial point is that in synchrony there is no diachrony. This is the problem. When a speaker produces a piece of parole (i.e. a spoken or a written text), he can only use the langue as it is at that moment. He cannot use his langue and change it at the same time. Nor can he use segments of text from previous moments because he cannot be in two different moments at the same time. Since texts are the most basic fact to investigate, it is disastrous that so many linguists have not understood that a text can only be synchronic. I have argued against some of those who

Göran Hammarström: Reply to Johannes Kabatek’s comment - 8 -

do not understand Saussure (Wartburg, Lehmann, Bailey, Malmberg, Chomsky/Halle, Martinet, Cermak and Fromkin) in Fundamentals of Diachronic Linguistics, München: Lincom Europa 2012, p. 12–14. Coseriu believes that he has shown that Saussure is wrong when he separates the two kinds of linguistics because language is changed when it is being used in speech. This is not an acceptable argument because change is not part of the speaker’s reality when he produces speech. Only an informed outside observer knows that language changes while being used and he may use this and other kinds of knowledge in diachronic considerations.

Kabatek disagrees only partly with Keller’s idea of using the invisible hand in order to understand problems in diachronic linguistics. I disagree completely. I see the most basic facts in diachrony as follows: A speaker sometimes invents something new in his language. It may be a sound, a word, a syntactic construction or something else. It has immediately among other functions idiolectal function. He may forget it again, in particular if nobody imitates him. It may also happen that the people around him start using the innovation. It then gets a sociolectal function with all its implications. (I have suggested the term sociolect for use in linguistics. A fact of language used by any kind of group of speakers and not by other groups is sociolectal.) If the innovation spreads to definite areas, it obtains dialectal function. And if it spreads to all speakers in the dialect, the sociolectal function disappears. (I have discussed the sociolectal and dialectal function of language in Zur soziolektalen und dialektalen Funktion der Sprache, Zeitschrift für Mundartforschung 34, 1967, p. 205–216 ). If the innovation spreads to the whole language, the dialectal function disappears. If the spreading of the innovation is less complete, it will have both sociolectal and dialectal function within the language (the glottolect). If one thus considers what happens in language change, I can only see that if one tries to use as an explanation the invisible hand, i.e. an event is the unintended cause of another event, one should see that it does not fit the description of language change. Language is changed by the speakers in the world of free will and nothing is caused. It is clear that when an innovation spreads to groups of speakers, it takes on sociolectal and dialectal functions that it does not have at the beginning. This can be and has been described and the details are interesting. The fact that the innovative speaker has more than one possible innovation to choose from is also proof that no cause is involved. One may wonder if the sociolectal and dialectal functions of an innovation which are not immediately intended by the innovator are caused. I think, however, that it is rather true to think that the spreading to given groups of speakers is not the cause of but the condition for these functions. The speakers provide intentionally the innovation with various kinds of these functions or leave the innovation rather neutral in this regard. Nothing causes the speakers to do one change rather than another. Adam Smith’s invisible hand certainly fits many problems in other sciences but it is inapplicable to problems of language because the speakers do not do things which without their intention cause language change. Kabatek finds that Keller’s idea that “the sum of individual finalities creates a ‘collective causality’” is problematic. I find it unacceptable because, as I have explained above, I cannot see that anything has been caused in the language.

When discussing language change, I believe that one should keep in mind that the interesting and often discussed problems mainly concern phonemes, morphemes, lexemes, lexeme sememes and syntagmemes and are of the kind which I have called major changes. When these are discussed, many notions such as cause, intention, economy and

Göran Hammarström: Reply to Johannes Kabatek’s comment - 9 -

analogy have been used in different ways by different linguists. There are also minor changes which are less problematic. They often have a strong causal element. For instance, Italian immigrants come to a country and the word pizza is added to the country’s language. Or some language authority gets an idea about a word and manages to get it accepted. I have also discussed diachronic problems in my Comment on Cristinel Munteanu’s contribution