Volume 3, Issue 1

1. Introduction A key intellectual advance in 20th-century linguistics lay in the realization that a typical human language allows the construction not just of a very large number of distinct utterances but actually of infinitely many distinct utterances. However, although languages came to be seen as non-finite systems in that respect, they were seen as bounded systems: any particular sequence of words, it was and is supposed, either is wellformed or is not, though infinitely many distinct sequences are each wellformed. I believe that the concept of “ungrammatical” or “ill-formed” word-sequences is a delusion, based on a false conception of the kind of thing a human language is.

1. Introduction Geoffrey Sampson defends the extraordinary claim that there is no theoretically significant non-quantitative linguistic difference between a sentence of English and a string composed of the same words in the opposite order. Let me put that a different way. Order opposite the in words same the of composed string a and English of sentence a between difference linguistic non-quantitative significant theoretically no is there that claim extraordinary the defends Sampson Geoffrey.

The ready availability of large corpora has opened a range of interesting possibilities for linguistic research, and Geoffrey Sampson's article illustrates that a corpus-based perspective can also motivate revisiting the general direction and methodology of linguistic research. At the same time, the discussion in Sampson's article pushes two issues to rather extreme conclusions, which I think could be useful to revisit here.

In my response to Sampson's Grammar without grammaticality , I will focus on two issues. First, I will develop an extended interpretation of Sampson's ‘Norwegian’ example, essentially following Sampson in arguing that there is no clear-cut distinction between grammatical and ungrammatical structures in any given language, but that, instead, there are simply more and less conventionalized structures. Second, I will discuss some methodological and theoretical consequences of this interpretation (or affirmation) of Sampson's arguments, arguing that traditional grammatical theory could be replaced by a general linguistic theory of the occurring and the non-occurring.

1. Introduction Sampson (this issue) argues for a concept of “realistic grammatical description” in which the distinction between grammatical and ungrammatical sentences is irrelevant. In this article I also argue for a concept of “realistic grammatical description” but one in which a binary distinction between grammatical and ungrammatical sentences is maintained. In distinguishing between the grammatical and ungrammatical, this kind of grammar differs from that proposed by Sampson, but it does share the important property that invented sentences have no role to play, either as positive or negative evidence.

1. Introduction As Geoffrey Sampson points out in his target article “Grammar Without Grammaticality”, a key concept of modern linguistics is the distinction of “good”, a. k. a. grammatical, and “bad”, a. k. a. ungrammatical, sentences. As such most linguists seem to subscribe to what I shall call the “Sheryl Crow view”, i. e., that grammaticality is a question of yes-or-no. Sampson, on the other hand, appears to take a more Hamlet-like approach in suggesting that “the concept of ‘ungrammatical’ or ‘ill-formed’ word sequences is a delusion” (p. 1). Instead he basically divides sentences into the “set of sequences which feel familiar to a speaker, and the set of sequences which are unfamiliar” (p. 11), with the latter including “sequences destined never to have a use, and those which will in due course be useful” (p. 11). In order to provide an adequate description of the set of familiar and unfamiliar sentences of a language, Sampson furthermore argues that linguists should only draw on corpus data, and not native speaker introspection.

Geoffrey Sampson has for quite some time now argued for an empirical basis of linguistics, as he has done in earlier works (e. g., Sampson 2001) as well as in this paper under discussion. It seems, however, that Sampson has hitherto considered only corpus-based evidence as sufficiently empirical, and he sees little or no role, not only for intuition, but also for experimentation, because he sees no difference with the researcher's own intuition and that of an (external) informant, characterizing this as “merely treat[ing] the informant's intuition rather the linguist's as the source of authority.” (2005: 28). In his present paper, it now seems that Sampson has fallen into a pessimistic stance not far from defeatism, when he asserts that not even corpora can provide reliable evidence on rare linguistic items and phenomena. And if one considers corpora as the only reliable source of empirical evidence in linguistics, and intuition as the only but in effect unreliable alternative, this is, of course, a conclusion that one could possibly reach. However, in our opinion Sampson's analysis is at times incoherent and often fallacious, and his view of the range of methods and sources of evidence available to linguistic research in general is too restricted, to say the least.

1. My thesis restated The responses to my target article are thoughtful and interesting. It is a matter of regret to me, though, that some of them say little about the issue which I aimed to raise.