Path: utzoo!attcan!uunet!lll-winken!lll-tis!helios.ee.lbl.gov!pasteur!ucbvax!CS.ROCHESTER.EDU!nl-kr-request From: nl-kr-request@CS.ROCHESTER.EDU (NL-KR Moderator Brad Miller) Newsgroups: comp.ai.nlang-know-rep Subject: NL-KR Digest Volume 5 No. 1 Message-ID: <8807112321.AA02499@castor.cs.rochester.edu> Date: 11 Jul 88 23:16:00 GMT Sender: daemon@ucbvax.BERKELEY.EDU Reply-To: nl-kr@CS.ROCHESTER.EDU Organization: University of Rochester, Department of Computer Science Lines: 512 Approved: nl-kr@cs.rochester.edu NL-KR Digest (7/11/88 19:13:18) Volume 5 Number 1 Today's Topics: looking for translator systems artificial languages Wanted: a grammar for English Re: Irregular forms [Was Re: Shallow Parsing] Re: Sapir-Whorf Hypothesis Submissions: NL-KR@CS.ROCHESTER.EDU Requests, policy: NL-KR-REQUEST@CS.ROCHESTER.EDU ---------------------------------------------------------------------- Date: Fri, 1 Jul 88 19:38 EDT From: UZR515%DBNRHRZ1.BITNET@CUNYVM.CUNY.EDU Subject: looking for translator systems H E L P ! ========= Dear subscriber, I am designing a Language Translator which should translate English Computer Text Books into persian. Indeed, everybody knows that there are a lot of works on automatic translation of "English-" texts into many different languages such as german, chineese, arabic, etc. It means that the "Analysing"-part of such a translator system is more than n times (naturally in different manners) designed and implemented. Now I will be very pleased to become information from any person interested in automatic translation of English texts (into any other language). Specially I wish to be informed where and how can I locate the followings: 1. Available translation systems from LISTSERV@FINHUTC or any other possibility to access such a system ? 2. A multi-lingua translator system such as EUROTRA ? 3. Fonts for persian or at least arabic characters ? Your prompt attention would be most appreciated. Yours sincerely, Hooshang MehrjerdianBergmeisterstueck 1 5300 Bonn 3 West Germany Tel: (0228)-733358 ------------------------------ Date: Wed, 6 Jul 88 01:02 EDT From: HEARNE@wwu.edu Subject: artificial languages I am doing a study of the design of artificial languages that might be accessible to computer support and would be grateful for any bibliographic hints anyone might have. I am particularly interested in whether anyone has ever written on the language LINCOS published by Freudenthal in the early sixties. If anyone is interested in the results of this inquiry, let me know and I will e-mail the assembled bibliography and/or suggestions. Jim Hearne, Computer Science Department, Western Washington University, Bellingham WA hearne@wwu.edu ------------------------------ Date: Mon, 11 Jul 88 09:37 EDT From: T. William Wells Subject: Wanted: a grammar for English I am looking for grammars for English. I'd prefer ones written as a context free rules with restrictions. I am currently looking through the references I picked up at the recent ACL conference and am investigating a number of other items I have picked up. However, I'd appreciate any references thrown my way. Please respond by E-mail, and if there is enough interest I will summarize to the net. ------------------------------ Date: Thu, 30 Jun 88 11:49 EDT From: Rick Wojcik Subject: Re: Irregular forms [Was Re: Shallow Parsing] In article <775@garth.UUCP> smryan@garth.UUCP (Steven Ryan) writes: >... It is hypothesised that languages change when someone >makes a mistake. If the mistake makes the language easier it spreads from >a one-time accident to a broad change. It is easier to believe this kind of thing when you are talking about the growth of vocabulary. But there seems to be a popular misconception that phonological change is initiated by imitation. Someone makes a 'mistake' and others join in. The first one to make such a mistake must be a pretty popular individual :-). The real driving mechanism of phonological change is to be found in language acquisition. Change spreads not only from a geographical center to a periphery, but from younger generations to older. An optional process in relaxed speech--e.g. the reduction of 'police' to [plis]--appears to children as the phonemic string /plis/. It is reasonable to suppose that /plis/ might some day become the 'standard' representation for the word. Since this type of phonological reduction happens to phonetic targets, rather than individual words, phonological change has global effects on the vocabulary. If change were driven by the imitation of mistakes, then we would see it as something that happened idiosyncratically, word-by-word, rather than across the board. >Possibly in early (?)Anglish, Saxon, or Protogermanic, somebody tripped >over their tongue when trying to say "hafed" and ended up with "had". The >irregular past of "have" (like most other irregulars) is essentially >unchanged from Old English (600 - 1100). The loss of fricatives between vowels and in the neighborhood of other consonants is a very reasonable phonological change. Consider the dialectal pronunciation of 'business' /bIznEs/ as [bInEs]. To understand the origin of 'had' you need to look at the pronunciation of other words with similar phonemic patterns in the centuries that the change took place. -- Rick Wojcik csnet: rwojcik@boeing.com uucp: uw-beaver!ssc-vax!bcsaic!rwojcik address: P.O. Box 24346, MS 7L-64, Seattle, WA 98124-0346 phone: 206-865-3844 ------------------------------ Date: Thu, 30 Jun 88 16:04 EDT From: Rick Wojcik Subject: Re: Irregular forms [Was Re: Shallow Parsing] In article <786@garth.UUCP> smryan@garth.UUCP (Steven Ryan) writes: >If irregular forms are suppressed and then resumed, they should reappear at >about the same time. If they are lost and then relearned, they should reappear >staggerred by their frequency. >Is there any evidence one way or the other? Let me try to restate what I said in different terms. Irregular forms don't exist in the earliest stages because there are no rules that they can be irregular with. So words like 'went' and 'looked' are morphologically simplex. When the morpho-syntactic rule for past tense is learned, it is only then that 'went' can be regarded (and must be learned) as an irregular form. One way to analyze the process is to say that the learner sets up an obligatory suppletive replacement between the regular form 'goed' /god/ and the suppletive form 'went' /wEnt/. So I am not saying that irregular forms are 'suppressed and then resumed'. In reality, there may be a period when the GOED->WENT suppletion process is optional. It is often the case in heavily inflected languages that learners go through periods of confusion (variable output) before they settle down to correct usage. -- Rick Wojcik csnet: rwojcik@boeing.com uucp: uw-beaver!ssc-vax!bcsaic!rwojcik address: P.O. Box 24346, MS 7L-64, Seattle, WA 98124-0346 phone: 206-865-3844 ------------------------------ Date: Fri, 1 Jul 88 12:56 EDT From: Rick Wojcik Subject: Re: Irregular forms [Was Re: Shallow Parsing] In article <1991@uhccux.uhcc.hawaii.edu> stampe@uhccux.uhcc.hawaii.edu (David Stampe) writes: [On idiosyncratic phonological changes in exceptional morphology, where th->h in 'I hink...'] There is also a variant of this where people voice the 'th'. It has the same distribution, viz. only in somewhat parenthetical uses of 'think'. I've just about bought David's claims about the etymology of 'had' ( Subject: Re: Sapir-Whorf Jim Meritt complains that in response to his request for evidence for or against the Sapir-Whorf hypothesis, he got nothing but a "one language is as good as any other" type reply. "Prejudice rears it ugly head - opinions without facts," he says. Well, some languages are better for rhyming than alliterating. Some are better for talking about fishing than farming. But Whorf was claiming something more: that grammatical categories of languages affect the perception and cognition of their speakers. The problem is, he didn't really show this at all. How could he have? There have always been many thousands of people who speak languages with totally different structures and grammatical categories. If they had observed time warps while switching tense systems, you can bet we'd have heard plenty about it. And among the completely divergent views of time held by various pre-Socratic philosophers, not one resembles the Greek tense system. The testimony of one amateur linguist based on a brief acquaintance with Hopi language, thought, and reality is not compelling, but if you're bent on the idea that some languages are more equal than others, that's the book to start from: the collection of Whorf's papers called _Language, Thought, and Reality_. The association of Sapir's name with Whorf's hypothesis always seemed to me to misrepresent Sapir's views. Sapir was a linguist with the sensibilities of a poet and musician, and therefore quite interested in the form of expression, and in classifying the variety of language structures and linguistic tendencies. Nonetheless, he often said (e.g. in his 1921 book _Language_) that each language is essentially complete and perfect. David (stampe@uhccux.uhcc.hawaii.edu) ------------------------------ Date: Sat, 2 Jul 88 15:52 EDT From: Sarge Gerbode Subject: Re: Sapir-Whorf Hypothesis In article <1031@aplcomm.UUCP> @aplvax.jhuapl.edu:jwm@stdc.jhuapl.edu (Jim Meritt) writes: > >I asked a month or so ago for any evidence for or against the S/W hypothesis. >(after a rather flamey "one language is as good as any other" here) > >I got a couple of emails asking for information that I received. The point is brought up in the book "Rationality and Relativism", Steven Lukes and Martin Hollis, Eds. (MIT Press, Cambridge, 1982) p. 267. He gives three references for studies that appeared to cast doubt on the hypothesis: Cole, M. and Scribner, S "Culture and Thought", Chapter Three (Wiley, New York, 1974) Heider, E.R. "Linguistic Relativity", in Silverstein, A. Ed. "Human Communication: Theoretical persp[ectives" (Laurence Eribaum Asociates, HIllside, N.J., 1974). Berlin, B., and Hay, R. "Basic Color Terms" (University of California Press, Berkeley, 1969). Lukes summaraizes the research by saying that it was found that "basic color terminology appears to be universal, and that the color space is 'naturally organized' into focal colors and that perceptually salient focal colors appear to form natural prototypes for the development of color terms." He adds that similar evidence has been found with respect to geometric forms and facial expressions of emotion: "both of these domains appear to be structured into 'natural categories' invariant across languages and cultures." References he gives for these last points are: R. Heider "Linguistic Relativity" and Clynes, M. "Sentics: the touch of Emotions" (Doubleday, New York, 1977), pp. 44-51. Why don't you check these out and post a summary, if the subject interests you. -- Sarge Gerbode -- UUCP: pyramid!thirdi!metapsy!sarge Institute for Research in Metapsychology 950 Guinda St. Palo Alto, CA 94301 ------------------------------ Date: Sat, 2 Jul 88 19:19 EDT From: morgan@clio.las.uiuc.edu Subject: Re: Sapir-Whorf I'm not sure what it means to ask whether Sapir/Whorf is correct. Nor how to go about looking for an answer with scientific merit. The same goes for the "one language is better than another" question. There is no evidence for the superiority of any one of the world's languages over any of the others. There can't be any such evidence, first of all because the question hasn't been stated clearly enough to be scientifically interesting, and even if it were there are no credible tools for measuring languages in relevant ways. History is full of pronouncements about the superiority or inferiority of particular languages. For example, I once heard a serious academic assert that Arabic is unfit for mathematics (a real howler). Such pronouncements are usually made by people who are looking for a justification for bigotry. That's why people with any sense at all avoid speculation that might be misused in ways they hadn't considered. ------------------------------ Date: Sun, 3 Jul 88 18:56 EDT From: Celso Alvarez Subject: Re: Sapir-Whorf Hypothesis In article <469@metapsy.UUCP> sarge@metapsy.UUCP (Sarge Gerbode) writes: >references for studies that appeared to cast doubt on the hypothesis: >Berlin, B., and Hay, R. "Basic Color Terms" (University of ^^^^^^^ California Press, Berkeley, 1969). The second author you are referring to must be Paul Kay, not "R. Hay". Paul Kay is currently a professor at the Linguistics Department, U.C. Berkeley. C. A. (sp299-ad@violet.berkeley.edu.UUCP) ------------------------------ Date: Mon, 4 Jul 88 06:47 EDT From: Celso Alvarez Subject: Re: Sapir-Whorf Thorough arguments against the Whorfian hypothesis are developed in Fishman (1960), "A systematization of the Whorfian hypothesis," _Behavioral_Science_ 5.4, 323-329. Fishman articulates his review of the effects of language on social behavior around the interaction of a double parameter: a) the data of language (lexical-semantic and grammatical), and b) the data of behavior (cognitive and social). On the basis of this framework, Fishman establishes four possible types of social reflexes of language: (1) reflexes of lexical-semantic structures in verbal behavior (thus addressing the question of *codifiability* of culture-specific notions); (2) reflexes of lexical-semantic structures in non-verbal social behavior; (3) concomitants between grammatical structures and cultural notions; and (4) concomitants between grammatical structures and social behavior. After a review of a number of experimental studies, Fishman concludes that there is not enough evidence to support in its entirety the Whorfian hypothesis beyond what constitutes a self-evident truth: that each language has its "character", its specific "personality" based on the ways in which it codifies lexically and it structures grammatically the community cultural notions. For Fishman, what in part determines cognitive processes is not the particular structure of a given language, but the very human capacity of language: "one might suspect that the impact of language _per_se_ in cognition and expression ought somehow to be greater and more fundamental than the impact of one or another language feature" (:336). A classic argument against the S/W hypothesis refers to the social and cognitive behavior of bilinguals. Are bilinguals socially and psychologically schizoid because they master two different languages? Conversely, if S/W were correct, "monolingual individuals speaking widely different languages should, therefore, differ with respect to their symbolically mediated behaviors" (:323). It is not clear that this is always the case. Haugen (1978, "Bilingualism, language contact and immigrant languages in the U.S. A research report, 1956-1970," in Fishman, ed. _Advances_in_the_Study_of_Societal _Multilingualism_, The Hague: Mouton) offers as a counterexample to S/W: (a) the coexistence of structurally diverging languages in the same speech community, where this linguistic duality does not necessarily entail a duality in bilinguals' social behavior; (b) the equivalence of a bilingual's cognitive processes in one or the other language: "The automaticity and comparative semantic emptiness of grammatical structures make it possible for bilinguals to master more than one language and guarantee that as a vehicle of thinking one language will perform as well as another" (:65); and (c) the existence of linguistic universals in a greater quantity than initially imagined. Some of the above points are, of course, debatable. But they stress that, at the core of the issue, is the question of *perception* and *conceptualization* of reality, and the *verbalization* of cognitive structures. Since experimentation on cognition is mediated by verbalization, I personally doubt that irrefutable evidence can be found, at the present stage of research, that linguistic structures *in all cases* condition cognitive processes or vice versa: "The time might (...) now be ripe for putting aside attempts at grossly 'proving' or 'disproving' the Whorfian hypothesis and, instead, focusing on attempts to delimit more sharply THE TYPES OF LINGUISTIC STRUCTURES AND THE TYPES OF NON-LINGUISTIC BEHAVIORS that do or do not show the Whorfian effect as well as the degree and the modifiability of this involvement when it does obtain" (Fishman 1960: 337; my emphasis). Celso Alvarez (sp299-ad@garnet.berkeley.edu.UUCP) ------------------------------ Date: Wed, 6 Jul 88 10:38 EDT From: Walter Rolandi Subject: Re: Sapir-Whorf Regarding the Sapir-Whorf debate: What IS the big deal about the Sapir-Whorf position? If it can be said that an implication of the hypothesis is that a person's thoughts are defined by the language that the person speaks, so what? What could be more obvious? In fact, this is a circular declaration. A person talks about the things that mediate reinforcement in a person's culture. The things that mediate reinforcement in a person's culture determine what a person talks about. Walter Rolandi rolandi@gollum.UUCP rolandi@ncrcae.Columbia.NCR.COM NCR Advanced Systems Development, Columbia, SC ------------------------------ Date: Wed, 6 Jul 88 16:12 EDT From: SIEMON Subject: Re: Sapir-Whorf In article <1036@aplcomm.UUCP>, jwm@stdc.jhuapl.edu (Jim Meritt) writes: > I would like people who jump up and yell "they all are just as good" and > "you can translate any of them to any other without loss" to either You suffer from a desire for absolute definitions. What the hell do you mean (or does anyone mean) by "as good"? No one claims that it is always possible to translate "without loss" -- it remains unclear what there is to lose in any given translation. > a. present information to back up their flame There were references in most of the repsonses to your note; even my own diatribe suggested some. The _Basic Color Terms_ book is especially good as the authors set out with the assumption that they could pin down here, if nowhere else, an influence of language on perception. They failed. (But the failure was interesting.) > b. shut up. > > after I provided an item that I thought was interesting which appeared > to back up the hypothesis. I seem to have missed your original. In any case the problem is not the data -- the problem is making any sensible, predicitive hypothesis out of Whorf's notions. > As to "how would you know?": How about > 1. Formulate a problem in 1st language. > 2. solve it > 3. translate formulation into 2nd language > 4. solve it in 2nd > 5. translate one solution into other language and compare What do you mean by "problem" "solution" and "compare"? Really. If you start from definitions dear to your own culture, and test by posing the "problems" to another culture, you are likely to get into a muddle about a. they didn't understand my problem b. i don't understand their answer wherein it is all too easy to dismiss someone else as ignorant or bizarre when the only problem is in the mechanics of translation. Suppose the translations turn out the same -- what does that prove? Suppose they turn out different -- what does THAT prove? Are you talking about mathematics or what? Even in math there are different, non-translatable solutions to the "same" problem. In a wider context it's unclear what a "problem" is, let alone how many "solutions" it has, if any. People operate within a culture; that culture includes a language. By and large the language will support the culture. So? Suppose, pace Whorf, that some people (Zunis, for example) structure time differently than Europeans do. Do you have any way to differentiate language as a source of this structuring from any other aspect of these people's human interaction? And do you have any possible way of showing that the differences in structuring imply an underlying difference in experience? If the Zuni know something about time that we poor English speakers don't (or vice versa), how would either of them know there was an unbridgeable gulf? Bilingual speakers don't seem to have PERCEPTUAL problems, though sometimes they get into confusions where vocabularies don't match up very well, or the syntax of the better known language affects discourse in the second language. -- Michael L. Siemon contracted to AT&T Bell Laboratories ihnp4!mhuxu!mls standard disclaimer ------------------------------ End of NL-KR Digest *******************