Wiktionary talk:About sign languages/Archive 1

This should probably link to the WT:BP conversation. --Connel MacKenzie 23:04, 13 September 2007 (UTC)

Unicode notations

Note in developing this policy page that there are at least two different ways English Wiktionary will use Unicode notations of sign language signs:

1. To describe the sign in detail, similar to how we use IPA to transcribe pronunciations of words in spoken languages.
2. To transliterate signs loosely but canonically to determine the appropriate title of each sign language entry.

Stokoe notation was once the only widely-known option for transcribing sign language signs in Unicode, but a more recent publication by Liddell and Johnson covers more detail of each sign and is not quite so biased toward ASL. Rod (A. Smith) 23:45, 13 September 2007 (UTC)

Index:American Sign Language (et al.)

There is definitely no consensus on how to include SL entries, and I, for one, think it's an intractable problem. But we have SL translations and can certainly have more. Can we list these at Index:American Sign Language (et al.), with a note like "Due to technical problems, there's no entry for these words, but they are listed as translations of English words."?—msh210℠ 18:08, 20 September 2007 (UTC)

Stokoe

Stokoe is problematic: it uses characters and forms that can't be used in a page title. ([], #) and super and subscripts. (Why people who design transcriptions of all kinds feel so compelled to use obscure or odd characters is beyond me!)

It would be very good if there was a usable transcription, but I don't know of one.

An idea might be to re-transcribe the Stokoe into a usable set, use that as the page title and then link it with the Stokoe shown as the text of the link. But I haven't really thought this through.

No one has a better system? You'd think someone would! Robert Ullmann 17:28, 1 November 2007 (UTC)

Rodasmith, above, q.v., mentions one by Liddell and Johnson.—msh210℠ 17:36, 1 November 2007 (UTC)

Why can't we use commonly accepted English translations for the sign as the title (Boy, Girl, Women, Dad)? I understand that would have problems and limitations but why can't we use that until we can figure out another system? --Zoohouse 19:51, 16 February 2008 (UTC)

In my opinion -- for what it's worth -- the problems outweigh the benefits.—msh210℠ 17:10, 19 February 2008 (UTC)

What would be the problems? --Zoohouse 19:39, 7 March 2008 (UTC)?

See below. Rod (A. Smith) 01:47, 6 July 2008 (UTC)

Suggested naming scheme

An important aspect of an entry title is that it allow a reader to look up an entry upon encountering the word. Imagine someone sees a sign made by holding the B shape near the chest, then circling the chest with it, ending in a brief hold at the chest. If we choose English translations as the titles of sign language entries, readers will be unable to look up a sign upon encountering it for the first time. If we use a notational system that encodes high-level elements of each sign in a consistent way, readers will be able to locate the entry for such a sign and determine its meaning. Systems like Sign Writing don't use Unicode, so they're out. The system by Liddell and Johnson appears to be rather verbose, too much to be suitable for entry titles. I suggest we modify Stokoe notation to be compatible with Mediawiki and easy to type on a keyboard. Following are the details of my suggestion:

• All signs shall be given titles of the form location-handshape-motion to represent the initial location, handshape, and motion (or orientation if no motion is involved) of a sign.

Since such a system is not terribly precise, there will frequently be multiple signs described by a given tab-dez-sig sequence. Those will be treated like homonyms, and each will get its own complete entry on the shared page for that tab-dez-sig.

In entry titles, the location of the dominant hand at beginning of the sign shall be given as one of the following:

Location symbol	Description
neutral	neutral location
whole-face	face, or whole head (symbol is superimposed ᴖ and ᴗ)
upper-face	forehead, brow, or upper face
mid-face	eyes, nose, or mid face
lower-face	lips, chin, or lower face
side-face	cheek, temple, ear, or side face
neck	neck
trunk	shoulders, chest, trunk
upper-arm	upper arm
forearm	elbow, forearm
inside-wrist	inside of wrist
back-wrist	back of wrist

In entry titles, the handshape of the dominant hand at the beginning of the sign shall be one of the following:

Handshape symbol	Description
A	fist (as ASL 'a', 's', or 't')
B	flat hand (as ASL 'b' or '4')
5	spread hand (as ASL '5')
C	cupped hand (as ASL 'c', or more open)
E	claw hand (as ASL 'e', or more clawlike)
F	okay hand (as ASL 'f'; thumb & index touch or cross)
G	pointing hand (as ASL 'g' 'd' or '1')
H	index + middle fingers together (as ASL 'h,' 'n' or 'u')
I	pinkie (as ASL 'i')
K	thumb touches middle finger of V (as ASL 'k' or 'p')
L	angle hand, thumb + index (as ASL 'l')
3	vehicle classifier hand, thumb + index + middle fingers (as ASL '3')
O	tapered hand, fingers curved over thumb (as ASL 'o' or 'm')
R	crossed fingers (as ASL 'r')
V	spread index + middle fingers (as ASL 'v')
W	thumb touches pinkie (as ASL 'w')
X	hook (as ASL 'x')
Y	horns (as ASL 'y', or as index + pinkie)
8	bent middle finger; may touch thumb (as ASL '8' in the latter case; this is a common allophone of Y)

In entry titles, the first motion (or orientation if there is no motion) of the sign shall be one of the following:

Motion/orientation Symbol	Description
up	moving, facing, or pointing upward (including "supinate", palm upward)
down	moving, facing, or pointing downward (including "pronate", palm downward)
up-down	moving up and down
dominant	moving or facing toward the dominant side
weak	moving or facing toward the center or non-dominant side
side-side	moving from side to side
to	moving or facing toward the signer
away	moving or facing away from signer
to-away	moving toward and from the signer
twist	twisting the wrist back & forth
bend	nod hand, bend or bent wrist
open	open up
close	closed
wriggle	wriggle fingers
circle	circle clockwise or counter-clockwise
together	approach, move together, near
touch	contact, touch, touching
link	link, grasp, or linked
cross	cross or crossed
in	enter, inside
separate	separate
exchange	exchange positions

I've used the above title encoding in the entry (deprecated template usage) trunk-B-circle. That entry obviously needs images, but hopefully it gets the point across. Notice how it should be relatively easy for a reader to locate an entry based only on seeing the sign, without any prior knowledge of its meaning. Rod (A. Smith) 01:47, 6 July 2008 (UTC)

Also note that the naming scheme is somewhat complementary to the indexing scheme rooted at Index:American Sign Language, since index pages like Index:American Sign Language/one-handed/B and Index:American Sign Language/two-handed/open A specify the number of hands and a more detailed handshape, while the entry naming scheme specifies where the sign is made and its initial movement/orientation. Rod (A. Smith) 16:43, 7 July 2008 (UTC)

Wow, thanks for "being bold", Rod. While this system is not perfect (obviously, in a perfect world, we'd have the signs themselves as the PAGENAME, which is impossible), it is as good as anything else suggested, and has the added benefit that someone is actually doing something about it.  ;-)  Might I just add that there is some talk in the SL community of getting SignWriting to be a Unicode block. I don't know how soon that may happen, but if and when it does, we may wish to move all these entries. For the mean while, though, I think it's good the way it is. I know tab-dez-sig is traditional; but are these specific values of those three parameters also traditional, or did you make them up?—msh210℠ 17:27, 7 July 2008 (UTC)

I didn't know about the initiative to add SignWriting to Unicode. I look forward to that, and yes, we should definitely move sign language entries to their SignWriting titles when Unicode, Mediawiki, and most web browsers support them. The specific values I chose above are my own invented English mnemonics of the symbols used in Stokoe notation. I'd have used one of the standard ASCII standards, but the ASCII standards I found use symbols that Mediawiki doesn't support in entry names ("[", "]", "<", ">", etc.). And, I figured, the audience for the English Wiktionary knows English already, so English words seems appropriate. I can't find any standard that uses English words, so I just chose common, short words that represent the general locations/movement/orientations. I certainly welcome any suggestions for shorter, more accurate, less ambiguous, or previously published values. Rod (A. Smith) 21:40, 7 July 2008 (UTC)

Signwriting has been added as a script code (ISO 15924) Sgnw. A proposed code block is in the "roadmap", starting at U+1D800 (e.g. on plane 1) but there is no proposed draft for the codepoints. This means it is probably a minimum of two years away, but OTOH can be expected to happen. Whether it will be usable in a linear representation remains to be seen. The copyright and control by Sutton is a serious problem; Sutton has not to my knowledge made any public domain or GFDL etc material available; and since it is a novel system, one can't claim it is free of copyright (as with, say, hieroglyphs or French ;-). (*sigh*) So for example, even though there are dictionaries on the site for free download, they are copyright, so one can't copy entries. Robert Ullmann 17:28, 8 July 2008 (UTC)

Valerie Sutton has released the latest symbol set called the International SignWriting Alphabet (ISWA) 2008 under the open font license. The ISWA 2008 is a double octet coded character set. The code pages are available. The proposed UTF code block has been available for several years but there is no active development for a solution using it. Steve Slevinski has a proof of concept for the SignWriting MediaWiki Plugin that will be released under the GPL3. We're getting there slowly. Regards, -Steve 19:14, 30 October 2008 (UTC)

Stupid question: does MediaWiki support non-BMP codepoints in entry titles? My impression was that it didn't. For example, U+1D800 is %E0%9D%A0%80 in URI-encoded UTF-8 (at least, I think it is — I did that manually, which is not how it's meant to be done ;-)), and http://en.wiktionary.org/wiki/%E0%9D%A0%80 redirects to http://en.wiktionary.org/wiki/%C3%A0%C2%A0%E2%82%AC (U+00E0, U+00A0, U+20AC — so apparently it's interpreting it as URI-encoded Windows-1252 and discarding the x9D). Though I suppose it's not worth worrying about, since even if it doesn't now, there's no reason to think that won't be fixed between now and when the codepoints are assigned, anyway. —Ruakh_TALK 18:34, 8 July 2008 (UTC)

Sure: 𒂍. or 𢄓. Full UTF-8 and ISO 10646-31. But the WP/Python code in 16 bit mode only supports up to U+10FFFF, however that is plenty for now. U+18D00 is F0 9D A0 80 in UTF-8: http://en.wiktionary.org/wiki/%F0%9D%A0%80 which works fine. (You lost a bit in the first octet. ;-) Robert Ullmann 23:56, 8 July 2008 (UTC)

Whoopsie, I'm so unused to non-BMP UTF-8 that apparently I can only count to three now. :-P   Unicode only goes up to U+10FFFF, and ISO 10646 is supposed to be synched with Unicode now, so in theory that should permanently be plenty. Sorry for the false alarm. :-)   —Ruakh_TALK 02:01, 9 July 2008 (UTC)

Another point: Your handshape names are very ASL-centric, which is not a good thing. This proposal is still better than what we've had hitherto, of course. (I know they come form Stokoe, but that was written for ASL, IINM.) (And, incidentally, I like "up-down" beter than "bounce", which latter gives an impression of sharp movements (down-up, then hold slightly, down-up then hold slightly) whereas the former has no such implication AFAICT.)—msh210℠ 16:04, 8 July 2008 (UTC)

I'm not sure what to do about the ASL bias. Do you know of a set of shape names that's less biased toward ASL? Regarding "bounce" etc., I may have taken the push to eliminate dashes too far. Given that there are no dashes in the first or second fields, any dashes in the last field are probably harmless. Robert, do you have any thoughts about whether dashes in the final "movement" field are to be avoided? Rod (A. Smith) 16:13, 8 July 2008 (UTC)

The way I see this is that there are two levels: people who know how to do simple things in code (find the first hyphen, then find another one), and people who know regex fairly well ([-a-z]+)-([A-Z0-9])-([-a-z]+) in the first case having tab be one word is helpful, in the second case it doesn't matter. In neither case do hyphens in sig matter. And I do concur that "up-down" is better than "bounce". Robert Ullmann 16:39, 8 July 2008 (UTC)

I don't know what to do abut the ASL bias, either. I was just bringing it up. I asked recently on a Usenet group devoted to linguistics and on a listserv devoted to sign-language linguistics whether there are language-neutral handshape names. I did not get any "yes" response.—msh210℠ 17:03, 8 July 2008 (UTC)

Following moved from WT:TR#a word meaning "to move to and fro":

I'm looking for a short, single English word to serve as a mnemonic for the act of moving to and fro, specifically forward and backward as opposed to side-to-side. (deprecated template usage) ping pong is out because it's two words. (deprecated template usage) tennis is the best I've been able to think of. Can anyone do any better? The word would replace "to-away" in the title naming scheme on Wiktionary:About sign languages. Rod (A. Smith) 05:51, 8 July 2008 (UTC)

How about (deprecated template usage) saw, as in the fore-back motion of (deprecated template usage) sawing? --EncycloPetey 07:35, 8 July 2008 (UTC)

Is there any ASL word that's (1) very common and (2) particularly characteristic of such motion, and whose meaning is such that the motion makes sense? If so, maybe go with its English translation? —Ruakh_TALK 17:04, 8 July 2008 (UTC)

Good ideas, both of you. Further discussion on Wiktionary talk:About sign languages, though, reveals that it isn't so important to eliminate the dash in the motion component of sign language entries. It seems we'll stay with -up-down rather than my hasty change to *-bounce, which had the unwanted connotation of a jerky motion. It's probably best to avoid ASL bias anyway, so since -up-down matches naturally with -up and -down, it now seems best to leave -to-away as is, matching -to and -away. (For what it's worth, an ASL bias applied to -saw might conjure up the sign (deprecated template usage) trunk-A-sides, which is made with the dominant hand in a fist moving side-to-side over the non-dominant hand held flat, as if the dominant hand holds the handle of a saw, cutting through wood represented by the other hand.) Thanks for your both of your suggestions, though. Rod (A. Smith) 17:40, 8 July 2008 (UTC)

Shall we use the word "wobble" "pendulate"

Unfortunately, neither of those words describes the direction of motion, and "wobble" would probably be read to mean the wrong motion anyway, one that we're describing as "wiggle". Further input is welcome, of course, but per Robert and Msh210, above, "to-away" is just fine and this turns out to be a non-problem after all. Rod (A. Smith) 06:24, 9 July 2008 (UTC)

How about pushpull as a compound word? Or row? --Joe Webster 06:02, 9 July 2008 (UTC)

"Pushpull" is pretty good, I'd say. It's a little longer than "to-away", but if others prefer it, I wouldn't oppose. Per Msh210 and Robert, above, though, there isn't really any need to eliminate the dash, so unless there's strong support for another option, I'll just leave it as it is. Rod (A. Smith) 06:24, 9 July 2008 (UTC)

It has more letters, but less syllables. Looking at the chart above, wouldn't pull-push be more accurate? --Joe Webster 22:20, 10 July 2008 (UTC)

In this broad sign language transcription system, the “to-away” value of the motion component means that there is a motion phoneme to the sign wherein dominant hand moves in two different directions: toward the signer and away from the signer. It purposely does not specify which motion occurs first, nor how many times the motion repeats. The reason it doesn't specify the order is that, for most signs that move both forward and backward the order is not phonemically contrastive (er, at least it's not in ASL, and I assume that rule applies to other languages). That is, if I sign (deprecated template usage) lips-O-close to-away lips-O-close in ASL, it doesn't really matter whether the first motion is toward or away from my lips, so long as I make the right shape in the right place, in the right orientation, and move my hand in both directions. Hope that makes sense. Rod (A. Smith) 22:57, 10 July 2008 (UTC)

Wiksign

I still don't think we should list signs under their glosses; I'm just noting that the online LSF wiki-dictionary (competition!) Wiksign does so. Of course, they don't have spoken-language entries under the same titles. They use MediaWiki, which makes navigation familiar to our users; check it out. Their content seems to be cc-by-sa, but I'm not sure whether that includes the images and videos; if so, then, as far as I can tell, we can use those images and videos on Commons and here.—msh210℠ 18:11, 14 April 2008 (UTC)

Production header

I presume this is placed exactly as Pronunciation is? Rod: You might want to add it to the last table in User:AutoFormat/Headers. Copy the table line for "Pronunciation", but add the NS flag. (It also has to be in "TOS" in the code, which I've done, put will not be effective unless it is in the table. TOS is the set of L3 headers that occur before POS, at "top of section" ;-)

If you end up with a different header, we can easily switch it. It seems just a little odd to me, but I can't quite place why (;-), some connotation I am thinking of? Doesn't matter now. Robert Ullmann 15:56, 7 July 2008 (UTC)

You presume right. Should I add the header to User:AutoFormat/Headers now or wait for further feedback about the somewhat arbitrary choice of "production"? The entries so far are still rough examples (missing such crucial elements as images, although it seems I might have a few native volunteers), so someone might find a better heading as we flesh them out. Rod (A. Smith) 16:29, 7 July 2008 (UTC)

Fairly harmless to add it. Easy to change again. Do poke me if you decide on something else so I can change the internal table (which I really ought to "export" to the control file, and read with the other flags ...) FYI as of 13 June, there were no L3 "Production" headers in the wikt; so it doesn't conflict with any other use floating around. Robert Ullmann 17:02, 7 July 2008 (UTC)

Whatever we end up with will need approval by vote, since it'll amend ELE, no?—msh210℠ 17:22, 7 July 2008 (UTC)

Yes. What happens is that we get a fairly serious draft of this page (the project page) and take a vote on adopting it. It would then specify that "Production" is used only in sign languages. (The header might be voted on separately if needed, but there will be other things, like the entry naming. ;-). For now, this is about not having AF flag it, and also having it understand that it nests with multiple etys under the same rules as "Pronunciation". — This unsigned comment was added by Robert Ullmann (talk • contribs).

OK. "Production" is now in AF's non-standard headers table. Rod (A. Smith) 21:40, 7 July 2008 (UTC)

Hyphens

Using the same punctuation within the symbols as to separate tab-dez-sig makes it harder to take them apart. Or may not be an issue at all, and the visual effect better.

Using hyphens which are common in English and other language increases the potential collisions. Which is, of course, no more terrible than any two languages, and the particular form (-A- in the middle) reduces it quite a bit. OTOH, these may be kind of confusing too. But might be thought about. Either function might be another character, including space. Just thinking about it Robert Ullmann 16:10, 7 July 2008 (UTC)

Good thoughts. When I realized that we'll include phrases like (deprecated template usage) trunk-A-supinate neutral-G-away, I chose hyphen so that space could separate consecutive signs, but I'm not committed to that decision. Perhaps spaces could be used within a sign, while comma followed by space could separate signs within a phrase? Open to suggestions. Rod (A. Smith) 16:29, 7 July 2008 (UTC)

I'm satisfied with the punctuation and spacing the way you did it, Rod.—msh210℠ 17:22, 7 July 2008 (UTC)

Indeed, it's not bad. Having signs be "words", e.g. not containing spaces and separated by spaces, is a good thing. (You both do realize that if this is at all successful, a lot of people may start writing this way? Even with homonyms, glosses or just context will serve to dab.)

One possibility is to note that the hyphens between tab-dez-sig aren't required, dez is UC or numeric, the others are lc; we might write "trunkAopen trunkGaway" (not as pretty though ;-). And/or tab and sig might use contractions. Just thoughts. Robert Ullmann 17:43, 7 July 2008 (UTC)

I wish there were a common, single English word for "upper arm", "back of wrist", and "inside of wrist". If there were, we could eliminate the dash in the Tab (location) component because the other dashy Tab values have relatively obvious dashless replacements. Then, the name would be easy to parse. I'll add a few more sample entries and then copy the above table to Wiktionary:About sign languages soon, knowing that we can still tweak it easily up until the policy approval vote. Rod (A. Smith) 21:40, 7 July 2008 (UTC)

Just upper, back, and inside might be just fine. Yes that makes them a shade more opaque, but not too badly. And "side-side" could just be "sides" maybe? I don't want this to be forced. (and I adore "wriggle" ;-). AF is running on the new config. Robert Ullmann 23:45, 7 July 2008 (UTC)

Thanks. I incorporated your suggestions, corrected "wriggle" to "wiggle" (-: or maybe we should keep "wriggle" since it's undeniably adorable ;-), and made a few other tweaks to the copy I added to the project page. Rod (A. Smith) 05:31, 8 July 2008 (UTC)

Example sentences

Based on the format for example sentences for Japanese entries, the example sentence for (deprecated template usage) B@RadialWrist-PalmAway Sidetoside is tentatively formatted as follows:

1. busy

   100px 100px 100px

   (deprecated template usage) Y@Jaw-PalmForward Y@Jaw-PalmForward (deprecated template usage) R@Chin R@Chin (deprecated template usage) B@RadialWrist-PalmForward-OpenB@CenterChesthigh-PalmDown Sidetoside

   “Yesterday, the restaurant was busy.”

That is, the first line of the example shows a sequence of images, some of which may be animated, then the second line shows something meant to correspond to the transliteration of a Japanese example sentence, then the third line shows a loose translation using natural English syntax. (Of course, it may make more sense when I get the images in there, but hopefully the intent is clear.) I'm not sure that's the best approach, so I hereby welcome any suggestions for improvement. Rod (A. Smith) 22:06, 9 July 2008 (UTC)

Hold-movement notation

Hold-movement notation is the system described in the Liddell and Johnson publication mentioned above in #Unicode notations. It is a successor to the traditional tab-dez-sig description used in the various versions of Stokoe notation, and now that I have a better understanding of it, I realize it's not necessarily as verbose as the first presentation I saw of it, so it's worth discussing whether we should adopt a variation of it for page titles.

Hold-movement notation breaks signs into something analogous to syllables (consisting of vowels and consonants) of spoken languages. Each hold can be described by a tab-dez-sig sequence using only orientation values in the sig component, and each movement can be described by one of the sig values for movement. Apparently, most individual words in ASL (and I presume in other sign languages), are of the form Hold-Movement-Hold. For example, the current entry (deprecated template usage) trunk-A-supinate could be encoded into hold-movement notation as trunk-A-touch supinate trunk-C-touch, which means, "hold dominant hand in A shape near the trunk touching non-dominant hand, then supinate the wrist, then hold dominant hand in C shape near the trunk touching non-dominant hand". As you can see, that's quite a bit more precise than "trunk-A-supinate". A phrase such as the one in the entry (deprecated template usage) trunk-A-supinate neutral-G-away would have a hold-movement notation like trunk-A-touch supinate trunk-C-touch diverge neutral-G-away. (Note: There are both advantages and disadvantages to the fact that one cannot necessarily determine where one sign ends and the next one begins in such phrase titles.)

This is still a somewhat low-resolution description, but it's quite a bit more precise than collapsing it down to a single tab-dez-sig for each sign. A very precise hold-movement description (like the ones I initially read) includes a finer division of hand shapes, details like continuity of the motion, extention of each digit, rotation of the thumb, non-manual signals, etc., and is normally presented in a table rather than a linear sequence. Low-resolution descriptions like trunk-A-touch supinate trunk-C-touch seem to me to be analogous to phonemic transcriptions of pronunciation, while high-resolution descriptions typically presented in tables seem analogous to narrow phonetic transcriptions.

So, it's now up to us to decide whether to use the very broad tab-dez-sig sequence in our entry titles or to make it a little more narrow and include holds and movements. Comments, questions? Rod (A. Smith) 16:59, 10 July 2008 (UTC)

So you're saying we would not typically have entry titles like Au.BAFIcBAFI.INMPBAVP.Au.BAFIcm0ST.INMPBAVP_..._B~u.ULcUL.BKHPBAVP.B~u.ULcm0ST.BAHPBAVP_._1o-p.cm0ST.TIFIVPMP? Because that's well nigh unreadable.—msh210℠ 16:56, 17 July 2008 (UTC)

As you can probably tell, I've been struggling to come up with a usable system. There are three competing forces at work: legibility, precision, and consistency with existing transcription systems. The tab-dez-sig system seems more readable at first glance, but is imprecise and could not be bot-converted to sign writing if we later have that option (and I would argue that it's only readable once you've studied the system for long enough, as is the hold-move system). I tried another variation of the hold-move system to make it readable, but the page names grew very long. Maybe that's not a big deal, though. I appreciate your feedback and agree that it's hard to read without learning the system. I will soon have a variety of examples for us to compare here, including one that minimizes length and one that maximizes both legibility and precision at the cost of inreased length. Bear with me, and keep your comments coming. Rod (A. Smith)

Following are candidate pagename titles for the short ASL phrase with the English translation "How are you?":

1. OpenA-OpenA-To-Supinate C-C-Up Neutral-1-Away
2. Au.BKFIcBKFI.INMPBAVP.Au.cm0ST.INMPBAVP ... B~u.ULcUL.BKHPBAVP.B~u.ULcm0ST.BKHPBAVP . 1o-p.cm0ST.TIFIVPMP
3. OpenA-Finger-InMplane-OpenA-Center-InMplane C-Ulnar-BackHplane-C-Center-BackHplane 1-Center-FingerVplane

The first candidate uses a modified version of Tab-dez-sig. It says this:

1. “An unspecified part of the dominant hand is located at or near an unspecified part the nondominant hand shaped as or similar to an Open A. The dominant hand is shaped as or similar to an Open A. An unspecified part of the dominant hand is facing toward the signer with an unspecified orientation. Supinate the dominant hand (and perhaps do something with the nondominant hand).”
2. “Then, an unspecified part of the dominant hand is located at or near an unspecified part of the nondominant hand shaped as or similar to a C. The dominant hand is shaped as or similar to a C. An unspecified part of the dominant hand is facing up with an unspecified orientation.”
3. “Then, the dominant hand is in an unspecified location in space. The dominant hand is shaped as or similar to a 1. An unspecified part of the dominant hand is facing away from the signer with an unspecified orientation.”

The second candidate uses precise hold-movement notation to say this:

1. “Hold the dominant hand with all finger pads contacting the palm and the thumb unopposed and open. The back of the fingers are in contact with the back of the fingers of the opposite hand. The inside of the hand is facing the midline plane and the base of the hand is oriented with the vertical plane. At the same time, hold the non-dominant with all finger pads contacting the palm and the thumb unopposed and open. The whole hand is contacting the location in space that is medially distant from the body, centered, and sternum height. The inside of the hand is facing the midline plane and the base of the hand is oriented with the vertical plane.”
2. “Then, move both hands straight to the next hold.”
3. “Then, hold the dominant hand with all four fingers together, mostly extended but lax (curved), and the thumb unopposed and open. The ulnar edge of the hand is in contact with the ulnar edge of the other hand. The back of the hand is facing the horizontal plane and the base of the hand is oriented with the vertical plane. At the same time, hold the nondominant hand with all four fingers together, mostly extended but lax (curved), and the thumb unopposed and open. The ulnar edge of the hand is in contact with the point in space that is medially distant from the body, centered, and at sternum height. The back of the hand is facing the horizontal plane below and the base of the hand is oriented with the vertical plane.”
4. “Then, stop using the nondominant hand and move the dominant hand straight to the next hold.”
5. “Hold the dominant hand with the index finger extended, the other fingers closed, the thumb opposed and its pad contacting the closed fingers on their radial edge. The whole hand is contacting the point in space medially distant from the body, centered, and at sternum height. The tip of the index finger is facing the vertical plane ahead and the whole hand is oriented with the midline plane.”

The third candidate uses a slightly less specific version of hold-movement notation (but still more precise than tab-dez-sig) to say this:

1. “Hold the dominant hand in the open A shape. An unspecified part of teh hand is near or in contact with fingers of the nondominant hand. The inside of the hand is facing the midline plane with an unspecified orientation. At the same time, hold the nondominant hand in the open A shape. The hand is at or near a point in space centered with the body at an unspecified height. The inside of the hand is facing the midline plane with an unspecified orientation.”
2. “Hold the dominant hand in the C shape. An unspecified part is near or in contact with the ulnar edge of the other hand. The back of the hand is facing the horizontal plane with an unspecified orientation.”
3. “Hold the dominant hand in the 1 shape. The hand is near a point in space centered with the body at an unspecified height. The finger is facing the vertical plane ahead and an unspecified orientation.”

Note that all three systems require the reader to be familiar with the details in order to interpret it accurately, but some seem easier to learn than others. Seeking feedback. Rod (A. Smith) 19:31, 19 July 2008 (UTC)

Handshape symbols

In some sections above, some concern has been expressed regarding the bias of using Stokoe symbols for the names of handshapes here. There are at least two ways to address this bias:

1. For each sign language, we could create a separate policy page that uses handshapes that are iconic for and phonemically contrastive in that language.
2. We can treat handshape names as we do IPA characters, which use a single set of latin (roman) characters and variations thereof to transcribe sounds from languages that are often not written using those characters.

It turns out that most of the handshape symbols used in Stokoe notation correspond well to the majority of fingerspelling alphabets. According to http://www.sign-lang.uni-hamburg.de/intersign/Workshop2/Miller/Miller.html:

Of the 44 countries or groups of countries surveyed in Carmel (1982), the vast majority share a significant number of handshape-character correspondences with the International and North American manual alphabets. (The principal differences between the two are that the North American handshapes corresponding to < F > and < T > are not used in the International alphabet since they are considered vulgar or obscene in many cultures)

Among the roman-based one-handed manual alphabets, those showing the most important differences in correspondence values are the Spanish-based alphabets, the Swedish and the Portuguese (from which the former is derived); these alphabets, nonetheless, still share a significant proportion of handshapes with values identical to those in the North American and International alphabets. Surprisingly, even two-handed alphabets include some handshapes that correspond to International one-handed values. Manual counterparts of non-roman scripts such as the Chinese, Japanese and Indian Nagari fingerspelling systems and the Israeli, Greek, Thai and Russian alphabets are based to a surprisingly large extent on roman alphabet values. The only cases of complete non-correspondence with roman values are fingerspelling systems based on the Arabic alphabet, the Ethiopian syllabary and the Korean Hangul syllabary-alphabet: in each case, handshapes are more or less iconic representations of the characters in the writing system.

So, I'm thinking it's not so bad that the handshape names are based on ASL letters. If any particular sign language project begins here and its participants would prefer an entry naming system that uses different handshape symbols, that should be fine so long as they are chosen to be phonemically contrastive within the language. If that happens, though, I think we should still have a production section in entries for that language that describes the sign in common, sign-language-neutral terms, analogous to how we use IPA in the pronunciation sections of spoken language entries. Rod (A. Smith) 17:03, 11 July 2008 (UTC)

Fwiw, I agree.—msh210℠ 17:57, 11 July 2008 (UTC)

strong and weak hands

I think they're usually called dominant and nondominant.—msh210℠ 17:12, 14 July 2008 (UTC)

Noted and corrected. Rod (A. Smith) 23:02, 14 July 2008 (UTC)

Fortunately, the entry pagename transcription system being proposed uses position rather than name to distinguish the dominant hand from the nondominant one, but for what it's worth, Padden and Perlmutter (1984) used the terms "strong" and "weak". It seems at least some other linguists are following that lead, but I don't know how exactly much uptake the new terms have enjoyed. Rod (A. Smith) 00:06, 16 July 2008 (UTC)