Talk:Zellig Harris

Untitled

In response to the wikify note, the article has been rewritten in sections. Is more needed? —Preceding unsigned comment added by Bn (talk • contribs) 16:51, 9 June 2008 (UTC)

Very informative and important article. Too bad it's written like an essay to the professor who is familiar with linguistics terms. It was tough going through the article, but worth it.

It seems odd to call him a Russian-American linguist when his family immigrated when he was four years old. He is a quintessential exemplar of American linguistics, in no respect is his linguistic work considered Russian or Russian-American, nor for that matter is Ukraine part of Russia.

Transformations and paraphrase

I've deleted the erroneous assertion that transformations as developed by Chomsky could only be justified by intuitions of paraphrase. For a start, Chomsky explicitly notes as early as Syntactic Structures that passive sentences are not always precise paraphrases of active sentences because of the effect of surface structure constituency on scope. He gives non-synonymous pairs like "Every linguist speaks two languages" and "Two languages are spoken by every linguist" to show this. What justified the passive transformation was not synonymy between sentences, but rather identity of selectional restrictions between active and passive forms of the same verb. In many other cases, untransformed structures bear no resemblance to English sentences, and it is impossible to have any intuitions about their meaning. For example, it is now commonly assumed that the underlying structure of a sentence like "John gave Bill the book" is something like "John v Bill gave the book", prior to a verb-raising transformation which derives the surface form. Clearly, the meanings of such abstract underlying structures are stipulated -- we do not have any intuitions about them.

The mistake of thinking that Chomsky's transformations are justified by intuitions of synonymy seems to be based on two misconceptions. First, that transformations transform sentences into other sentences. Since the input to a transformation is always an abstract structure, it is impossible to have intuitions of synonymy, since we clearly have no intuitions about the meanings of abstract syntax trees. Second, that base structures are the sole input to interpretation (i.e. generative semantics in one form or another). Chomsky has never assumed this. Cadr 20:27, 21 March 2007 (UTC)

Harris transferred the concept and term transformation directly from linear algebra, where transformations are mappings from subset to subset in a set. Abstract syntactic trees have no place here. Harris's formal criteria for transformation are not dependent upon intuition, and establish these mappings in the set of sentences of a language. In Operator Grammar the verb give is an operator that takes three zero-level arguments. No abstract structure is required, and no rules operating on such a structure. Linearization establishes the linear order of each operator and its arguments; reductions determine the morphophonemic shape of the operator and argument words, including zero allomorphs. [User:bn]] 16:47, 09 June 2008 (UTC) —Preceding unsigned comment added by Bn (talk • contribs)

Recent revisions

The article as I found it had some useful material, but a number of serious problems, of which I've tried to address only the easier ones. First, the article was (and to some extent still is) an encomium rather than a factual report, with too many assertions of its subject's profundity and greatness, unsupported by verifiable citation (e.g. "Harris's enduring stature derives from the remarkable unity of purpose which characterizes his work."). Second, the article (including some of the encomiumizing) appears to be pushing a somewhat obscure but not unfamiliar agenda: that Zellig Harris, not Noam Chomsky, had the key ideas responsible for the revolution in linguistics known as generative grammar. No facts known to me support this, besides the fact that Harris was Chomsky's teacher, encouraged him in his earliest work, had a theory of syntactic analysis that invoked a concept called "transformation", and was reportedly quite bitter about Chomsky in later life (reportedly refusing to set foot in the Penn linguistics department ever again after they hired a generative linguist). It is just not true that a book of Harris's "includes the first formulation of generative grammar", unless we interpret that term so broadly and fuzzily that it has no meaning. So this stuff has to go, and I've made a first step towards deleting it.

Now the idea that Harris was the real first generative grammarian (and that Chomsky ripped him off) is not unfamiliar to me, as I noted above. As a linguist, I have met people who know someone who has a friend with a cousin who made such an allegation. But to my knowledge it is not a supportable or verifiable claim, which is probably why the Wikipedia article resorted to the following formulation, with no citation: "Because of this, some linguists and historians have questioned whether Chomsky's transformational grammar is as revolutionary as it has been usually considered." It's also probably why the claim that Harris was developing generative grammar early in his career was sourced in the previous version to a private e-mail between two third parties (more or less the epitome of non-verifiability). I have deleted these passages, but of course if there are serious sources that could be brought to bear on the question, they could be cited in a future version.

There are other changes that could be useful, which I have not attempted. For example, a genuine contribution of Harris to later linguisticsis the fact that he was one of the first American structuralists to work intensively on syntax (though here too others such as Rulon Wells are important). Probably the article should highlight this fact more. (I have not made such a change.)

Some of the technical material in the article is obscure, and needs clarification as well.LanguageGreek (talk) 22:36, 22 July 2010 (UTC)

Yes, it did lack citations. I've added some, including all those tagged. Articles can always use more copy editing and work on tone. Please familiarize yourself with the literature and discuss here before deleting. Bn (talk) 21:04, 28 July 2010 (UTC)

Too technical

For non-linguists, this article will be rough going. Can somebody simplify it, or at least add in explanatory remarks where things get technical? Jaylgordon (talk) 02:03, 4 June 2016 (UTC)

Distributional semantics and distributionalism

I've found many citations of "distributional semantics" but not of "distributionalism". Z.S. Harris played a major role in spreading Distributionalism. I'm creating this page, please help.--S.vecchiato (talk) 12:17, 29 January 2020 (UTC)

References