Talk:American Sign Language grammar

From Wikipedia, the free encyclopedia

Articles for deletion This article was nominated for deletion on October 2007. The result of the discussion was no consensus..

why the frick is every sentence here OSV? That's not even the default word order of ASL! 71.58.168.204 07:54, 1 June 2007 (UTC)

It's not OSV; it's topic-comment, and the object is often the topic of the sentence. LordRM 20:28, 31 July 2007 (UTC)

Contents

[edit] Encyclopedic?

This is the least encyclopedic article I've read on Wikipedia. I don't know anything about the subject matter, so I don't think I'm qualified to fix it, and I don't really know how to tag it to point out this fact, but something needs to be done. It's informative, yes, but not nearly up to encyclopedic standards. --ZxqamF 03:19, 24 September 2007 (UTC)

Almost completely replaced with the material from the main article, though a few statements and some of the examples have been retained. (You can tell the old material because I didn't regularize the formating.) kwami (talk) 22:58, 17 May 2008 (UTC)

[edit] word order

ASL is strongly AVO, not OAV as the article now has it. Unfortunately, I won't have a chance to fix this up until next year. Could someone else do it? kwami 16:53, 31 August 2006 (UTC)

Okay, got it, along with a fairly extensive rewrite. However, many of the links are bad, or don't direct to the proper articles, and I don't have a good enough connection to sort it out. Sorry to leave someone else with the mess! kwami 13:58, 12 September 2006 (UTC)

The confusion about the word order I think comes from the large number of topic phrases and tags used in ASL. One dictionary I consulted even uses OSV as the normal order! (The American Sign Language Handshape Dictionary, Tennent and Brown, 1998) FoiledAgain 00:08, 14 December 2006 (UTC)

I have seen Object-Subject-Verb as the main structure for word order in multiple books including text books, and in use. BOOK+(topic) I LIKE; (I like books.) BOY HE(topic) GIRL HIT; (The girl hit the boy.) COFFEE BLACK(topic) I HATE; (I hate black coffee.) It is by no means the only sentence type in ASL, but it is a prevalent one. Some of the examples from the article are incorrect. "The dog chased my cat" could be signed CAT(topic) DOG MY CHASE. It could be signed CAT CL:v DOG CL:v (chase) using classifiers to show who chased whom. I would almost never sign DOG CHASE CAT. Notice in each of these examples, the verb is last.... Cwterp 15:09, 17 May 2007 (UTC)

'Topicalizing'? JoeSmack Talk 16:27, 17 May 2007 (UTC)
'Topic/Comment' is how the standard word order is normally referred to. But the topic is not always the object, it can be the subject too. Cwterp 18:30, 17 May 2007 (UTC)

[edit] handshapes

Quoting from the handshapes section of this article: "These handshapes are constrained in their interactions. For example, the 5 and F handshapes only make contact with another part of the body through the tip of the thumb, whereas the K and Y/8 handshapes only make contact through the tip of the middle finger, and the X handshape with the flexed joint of the index finger. The L hand always makes contact by means of the thumb" I can think of several exceptions off the top of my head to this "rule"... specifically for 5, F, and L. Any of you linguists able to explain this to me? How is "contact only made" even though there are exceptions? Cwterp 15:03, 28 May 2007 (UTC)

I can also think of exceptions. I think tends would be a better word here, although I have seen a room full of people bust up when a new signer used a silly point of contact with a handshape. It might be worth noting that other sign languages have different points of contact too.
As you can see this article stands to be improved; I'm glad you're going through it! :) JoeSmack Talk 15:40, 28 May 2007 (UTC)

[edit] Allophonic (or not) handshapes

Just edited this.

Previous version claimed that these were allophonic groups. Here are some minimal pairs (or near minimal pairs) that prove otherwise; I'm sure others can think of more:

  1. A, S, T
    PRIVATE vs ICE CREAM vs [nonexistent?]
    can't offhand think of any T signs that aren't initialisms
  2. B, 4
    WINDOW vs LIST
  3. 1, D
    MEET vs DATE
  4. Y, open-8, horns, ILY
    WRONG vs PREFER vs [nonexistent] vs [nonexistent]
    THAT or YEAH vs AWWW vs BULLSHIT vs I-LOVE-YOU

No source was cited for the previous list, so I've edited it directly (and also edited for style uniformity). If there was a source, feel free to change it to "X said that it's blah, but they're wrong because of foo".

I would suggest someone get handshape diagrams a la Klima & Bellugi _Signs of Language_ that show true allophones (e.g. B with thumb in front vs on side; E with back of hand straight & fingers not touching vs with back curled and fingers barely touching; horns with fist vs lotus hand; Y with fingers tucked vs held close vs held close w/ ring finger up due to physiological contraints), etc.

Also, the touching constraint section is inaccurate; e.g. F can touch not just at thumb tip, but with flat of the O part (e.g. POLKADOT) ; L can make contact w/ back of hand (LOSER), etc... so I've edited those less exclusive statements too. Sai Emrys ¿? ✍ 00:42, 9 December 2007 (UTC)

This was per Stokoe. While he didn't deny that there was variation, he found that it was not phonemically significant, excepting numerals and initialisms.
You haven't provided any minimal pairs except for initialisms, unless it's the last one (I'm not familiar with 'awww' or 'bullshit').
  • 'Date' is an initialism.
  • 'Window' and 'list' are both made with the B hand; the location and movement are different, not the hand shape.
  • In 'private' vs. 'icecream', Stokoe would argue that the phonemic distinction is one of movement and location: tapping the nose/central face vs. stroking the mouth/lower face, not the position of the thumb.
  • 'Wrong' vs. 'prefer' may be a good pair, though it's not a minimal pair, as the movement is also different. Stokoe did use different symbols, 8 vs. Y, for sets like these, but was never able to find words where the distinction actually depended on whether you sign 8 or Y. There was always something else going on, as in your example.
  • 'That' and 'yeah' use different hand shapes, Y vs. A, wheras 'I love you' is both an initialism and has a very different movement and orientation.
As for point of contact, 'Loser' is an initialism (it's also a borrowing from English and more recent than Stokoe's studies). I don't know about 'polkadots'; Stokoe might have argued that it's a diminutive of 'dots' with the C hand shape; he might have transcribed it as a closed C rather than F. (I'd have to check, and I don't have his stuff handy.) Can you think of any independent sign where an F hand touches with anything but the thumb & index?
Stokoe was able to find minimal pairs for all 17 hand shape phonemes he posited, as well as for location and movement. He noted that there were subphonemic differences, such as touching the chin for some signs, and the lips for others. However, no two signs are distinguished by whether they touch the chin or lips, and in fluent signing the difference gets blurred. Therefore he argued that 'lower face' is the phoneme, and that 'chin' and 'lips' are allophones. The same is true for 8 vs. Y.
Unless you can demonstrate something more concrete, I'd be tempted to say this is OR. kwami (talk) 21:40, 10 December 2007 (UTC)
DATE is indeed an initialism. But cf. DINNER (D) vs [nonexistent] (1). Clearly even though this is an initialism, if DINNER can be signed with a D but not with 1 (or at least, is wrong with a 1), the difference is significant.
TTBOMK, WINDOW can only be signed with a fingers-together B hand, whereas LIST can only be signed with a fingers-apart 5. WINDOW in citation form does have a dual movement whereas LIST has single, but e.g. OPEN_WINDOW will have about the same movement as LIST.
PRIVATE vs ICECREAM is, again, not a perfect minimal pair. However, PRIVATE with an S handshape would be just wrong (or maybe taken as an initialism for SECRET?), and ICECREAM with an A handshape would also be wrong.
WRONG and PREFER in my usage is a single short motion towards and hold at chin. Both have other forms - e.g. WRONG with a side-to-side roll, PREFER with multitap, etc. But finding any forms that overlap makes for a phonological difference. And PREFER can't be signed with proper 8 (middle & thumb touching), only an open 8 (= spread 5 w/ bent middle).
Another pair for Y vs open-8 is MEDICINE vs an alternate form of THAT. Again, not a perfect minimal pair since MEDICINE has a little wiggle and doesn't need the approach movement, but still, you can't swap the handshapes and have it be recognizable.
YEAH is an initialism of YES (normally S, initialized to Y). Again, not the best example perhaps, but the point is that it's definitely NOT interpretable as BULLSHIT. (These are all minimal movements in neutral space for me btw.) ILY is to me not really an initialism because it's so ASL-ized at this point that it's essentially a ASL-only gesture.
So, re phonology, I would say that the exclusion of initialisms is just analytically invalid; the question in phonology vs phonetics is, simply, do speakers distinguish between the two forms or does it merely sound weird or accented? And clearly the answer here is that yes they are distinguished. It's not always easy to find exact minimal pairs, and I grant I haven't done so here, but you CAN swap features (like I described with PRIVATE vs ICECREAM) to test this, and see if the swapped version would be recognized as correct or accented-correct. It's not. Ergo, they're different phonemes.
 
CHIN and LIPS is a quasi-tautological minimal pair for those locations. So is e.g. MOUSTACHE vs THIN_BEARD. But I can't think of any non-body-references that distinguish those two areas. However, I can't think of anything where Y and open-8 as being interchangeable; can you give some examples of where that is the case?
I can't think of any cases where F touches with anything but either the tip of the thumb & index or the outside flat surface of the mini-O formed by them. My edit reflected this - usually they touch at the tip but sometimes at the outside. That this may?? only occur as a reduction of C doesn't change that fact, because if it did, then C would reduce to F and F would still be constrained to touch at the tip only... which it doesn't. (Just like C can touch with the thumb tip, all fingers' tips, non-thumb tips, or outside flat of C...)
Don't get me wrong, I've read and liked Stokoe's work, but if he claims that something is allophonic when one can find examples where it isn't, then I would consider that correction to be self-evidencing and not an example of OR. I'm not claiming anything that one needs my authority on; I'm just referring to things you can look up in your ASL dictionary or native speaker brain of preference, the usual methods for determining allophonicity, etc - rather than coming up with novel, undocumented claims of fact (like, say, if I were to claim that I have seen PREFER signed on the nose).
 
I think that the "can you swap the handshapes" test is one of the simplest to do for a complex phonology like ASL's, and will show that the things I've distinguished are not allophonic, despite any claim by Stokoe that they are.
 
My friend Alex Fink reminded me of one other point that you have raised but not addressed directly is that ASL, unlike any natural spoken language I know of, has a phonology with variables. E.g. STACK_OF_PLATES can be anything from a collapsed G to a large C, with the exact degree of size being clearly phonemic. This is a more visual reference, perhaps, but that doesn't disqualify it on the simple criterion of "is it distinguished". You gave another example, with POLKADOT of large or small dots. Another would be SMART (with C handshape) which is only metaphorically rather than directly visual - another good indicator that this is grammatically relevant rather than "merely mime". Another, not with handshape but still phonological (on movement / path), would be OLD vs EXTREMELY_OLD, analogous to English /old/ vs /o:::ld/. Etc etc, I'm sure you can think of more examples.
My point here (totally separate from the above re allophones & contact points) is that we should probably add some section that describes this phenomenon, and perhaps gives examples of how handshapes can and can't vary.
 
Hope that helps and sorry for the tl;dr. ;)
ETA: AWW is a indexical version of FEEL - tip of bent open-8 middle finger aims at person being sympathized with (including possibly self); various forms use repeated 'touch' or elliptical motion. Not sure whether there's a standard gloss for this, but AWW is the closest I can think of. Sai Emrys ¿? ✍ 01:58, 12 December 2007 (UTC)
I think the new section you suggest is a good idea. Also, Stokoe's attempt at determining phonemes was in part motivated to show that ASL is a real language and not some kind of gesture gibberish. That was a real issue in the 1960s, but not one that needs to direct our attention today. Stokoe was the first to address ASL linguistically, and he was in uncharted territory, so maybe he wanted to keep things nice and concrete. Minimal pair = phonemes; no minimal pair = sub-phonemic distinction. There are a couple considerations here: (1) perhaps not all ASL phonemes have minimal pairs to distinguish them. Some dialects of English might not have any pairs for /ʃ/ vs. /ʒ/, for example. (2) Perhaps cheiremes do not have the same phonemic behavior as oral phonemes. Stokoe would have denied that, but I don't know if anyone knows for sure.
Initialisms are of course part of ASL. However, do we wish to consider /x/ and nasal vowels English phonemes because they occur in Bach and vin blanc? I think that was Stokoe's point: If you exclude initialisms and other English borrowings, what are the phonemic distinctions of "pure" ASL?
However, just swapping handshapes to see if the result is intelligible is not a good test for phonemicity. If you pronounced take without aspiration or release, people will mistake it for day (I know: my mother couldn't understand me after my first trip abroad), but that doesn't mean aspiration and release are phonemic in English, or at least it begs the question of what is phonemic, and if there is such thing as a phoneme. Another example: every has 2 syllables, mammary has 3, while memory may have either 2 or 3. I think this may have been the kind of difference Stokoe saw between, say, the A and S hands.
Maybe we could separate 'basic' phonemes which distinguish minimal non-initialism pairs; phonemes that only occur in initialisms (which of course might migrate into the first category); things which only occur in derivations, such as the F which contacts like a C in 'polkadots'; and things which only occur in mimesis (like 'chin' vs. lips'; perhaps 'polkadots' belongs here). Mimesis and interjections aren't usually considered in English, or we'd have to conclude English is a click language, but then mimesis is much more central to ASL, so I don't know. I'd like to see a professional reference, though, and not just what we think it ought to be.
kwami (talk) 05:11, 12 December 2007 (UTC)


I know the background you refer to, and I think that mentioning this fact as historical context (possibly even historical context specifically to the linguistic analysis, in addition to generally to the language per se) is important. Similarly I've seen what I more modern things I would characterize as backlash to that, claiming that ASL is totally separate from English, which is also clearly not the case for something that shares a spelling system, lots of words directly taken, productive initialisms, etc. (Can't come up with a good reference for something that does this though. :|)
I would believe that 100% discrete phonology is a significantly mistaken theoretical conceit when applied to ASL, which has numerous instances of grammatically relevant, non-discrerte phonological gradation. This in itself is quite remarkable, at least as a feature of spoken vs signed languages generally, of which ASL is TTBOMK by far the most studied.
I believe that /x/ MUST be considered at least marginally part of English, because for instance we do NOT import Mandarin tones, Russian /s\/, etc - so we should make a distinction. (Even though /x/ does tend towards /k/, it's not an instant change whereas the others are.)
I find this claim of a distinguishably "pure" ASL to be even less tenable than your analogous one in English; whereas /x/ is extremely rare, nonproductive and gradually being converted into more normal phonlogy, initialisms are numerous and spontaneously productive, and increasing rather than decreasing in number. I would suggest that a better way to analyze it would be simply to characterize better ASL's historical sign change tendencies (symmetry, dominance, centrality, etc). But I readily admit that it's a weird case, especially when we also consider the presence of SEE as a sometimes-contrastive sometimes-integrated bastard twin, and the therefore sometimes (!) denigrated social status of initialisms and SEE-isms in ASL as not being "deaf enough" etc etc. This in itself is IMO very much worthy of discussion and description - but I do not consider it a valid basis on which to say "no this is not part of the language" when clearly people DO use it, and not just in a special and declining set, as I mentioned.
 
Your suggestion for separating out different phonemes is I think a good one. One minor objection: F is not purely a derivative of C (cf CAT) - perhaps you meant this only as a touching constraint, i.e. "F may only contact at the tips of thumb & index, unless it is a diminuitive of C, in which case it may also contact at the outside flat of the circle"? I'd agree with that, as I can't think of any counterexamples. It would, again, be a somewhat unusual phonotactic...
Surely interjections are "real English", and |tsk tsk| (/!\!\/) is English as well - we can just say that clicks only occur in extremely limited, unproductive contexts in English. Are you suggesting that there are similarly phonemes in ASL that occur only in extremely limited, unproductive contexts?
 
I've thought some more about S vs A, and there are certainly some signs in which they ARE quasi-interchangable or close to it - e.g. PRESIDENT and SHOES... but for all of these I cannot think of a one where the 'correct' form is not S. And still, PRIVATE does not take S. The only alternate analysis I can think of is that A and S are allophonic but contextually constrained, e.g. if you were to say that the A form only occurs at the chin and S elsewhere. But again that'd be false - viz. ASSHOLE as A in neutral space vs YES as S also in neutral space (ish).... so while I can't come up with a perfect minimal pair, they don't seem to me to be in predictable variation or complementary distribution, nor are they interchangably understood in all signs - just in some.
My understanding at this point is that we agree T, D, and ILY are initialism-only phonemes*; A and S is unclear; B and 4 are contrastive per my example (another: GRAY vs WHATEVER); and Y, open-8, and horns are contrastive. Is that accurate?
If yes, I suggest we implement your suggestion and segregate phonology by alphabet-only** (eg Z), alphabet or initialism only (eg T), productive alphabetic phonemes (eg F), and productive nonalphabetic phonemes (eg open-8).
 
* ILY is also used in one form of MAKE-LOVE as an initialism?? off of BE-TOGETHER (= TOGETHER with durative cyclical motion). But can't think of anything else, so it seems to be (at least for now) a unique handshape analogous to English clicks, used slightly more productively but still very limited.
** Z is used quasi-productively in PIZZA with a double-Z as fist w/ bent index & middle finger (dunno if it's regional). This is also the SNAKE classifier handshape. Just noticed that it's not listed, so I've added it. Should we consider SIT allophonic to this? Again, that would seem quite weird to me - and the SIT handshape is also not represented here...
 
Given Stokoe's preeminence - or at least seminality - in this field I suggest that we mention what he thought vs what we are giving as at least the modern case, and perhaps also mention the limits and novelty of his analysis when it was published. Do you know if he revised it in later publications? What's the date of authorship of the source from which you cite?
Sai Emrys ¿? ✍ 10:24, 13 December 2007 (UTC)
His Dictionary was I believe 1965. I don't know if he kept up with later developments. I'm sure the work's been done - UC San Diego, maybe? - but have never seen it.
Oh, make love. I misread it as "I-love-you". kwami (talk) 21:36, 17 December 2007 (UTC)

[edit] ASL grammatical structure?

Can someone who knows ASL comment on the structure of sentences like ITALIAN ME ("I am Italian") — which was mentioned in one example in the article but not singled out in any way for discussion? Specifically, I'm wondering about the sign order here. Do "(noun/pronoun) — (am/are/is) — (adjective)" sentences in general show this same structure, with the adjective first and the subject last? I wonder if it might suggest that the subject is actually taking on a verb-like character here (i.e., that the ME sign in ITALIAN ME might really mean something more like "I am") — something which, if true, could make this structure consistent with the generally verb-final character of ASL sentences. Has anything been written about this that might be citable in the article as a source? Richwales (talk) 04:27, 21 April 2008 (UTC)

It has to do with information flow, what is topical vs. new, etc. Too much to go into here, and the necessary studies might not have even been done. kwami (talk) 00:13, 18 May 2008 (UTC)