User talk:Positivesigner

This user is reaching out to ASL contributors like Rod A. Smith for insight on how his newly developed ASL transcription code might be of use to the Wiktionary project. Samples are located at http://aslsj.com and http://aslsj.blogspot.com/. - Tom 08:36, 25 March 2009 (UTC)Reply

Hi, Tom. I replied briefly on my talk page to your request for comment, but I'll try to be a bit more detailed here. Your system is interesting. It would be useful to compare your list of phonetic parameters with those currently in use here. I'd be very curious to see where the two systems overlap and, more interestingly, where they don't. The primary focus here is mostly at the lexical level than at the discourse level, so the system here omits the harder-to-quantify non-manual-signals (e.g. looking left or right). Note, this omission is intentional, despite the fact that some non-manual-signals are crucial to certain lexical signs (e.g. the tongue protrusion, head shake, and head tilt in NOT-YET). Nevertheless, it could be useful to include such signals in transcriptions of example sentences. I'm really excited to confer further with you! —Rod (A. Smith) 22:11, 29 March 2009 (UTC)Reply

Actions speak louder than words... or signs... except that sign language words are actions... Where was I going with this?

Oh, yeah. I have yet only seen a few other contrived ASL transcription methods that use ASCII characters, most meant for signed languages in general, and they were very terse. Stokoe notation and Sign Writing are the only symbolic transcription method I've seen, again meant for general signed languages. Mine is unique to my own situation.

I happen to learn better from written material than visual - ergo my immediate conundrum in learning ASL. As a computer programmer, I have actually learned many contrived languages used to communicate program structure. I began to notice patterns in ASL production long before I actually understood the meanings. In order for me even to remember what signs were made on a DVD video, I had to create a to write it down. Once meanings were associated with the patterns, my code really started to turn into a whole transcription system.

I am comparing what I went through to produce my ASL Sign Jotting code with Louis Braille. (http://en.wikipedia.orghttps://dictious.com/en/Louis_Braille) The raised letter impressions that blind frenchmen were accustomed to read were hard to use and impractical produce. Thus those blind could read but never learned to write. He invented his own system of transcription that became the standard of raised notation, but only after his death. I am rather hoping my system won't take that long.

I am capable of comparing two transcriptional systems. However, like comparing raised letters to a Braille code, there is little in common. Mine is not meant to convey accuracy of production (what the sign looks like), but to differentiate meanings of similar signs. English can spell similar sounding words differently (cache and cash) as well as sound out similar spelling words differently (action do and musical do).

I have removed most of the phonetic parameters from my system because the bulk confused meaning. Approximation is something true signers do a lot in real-time singing. I transcribe the approximate sign and assign a meaning. Several closely-jotted signs, perhaps different on only one axis, may be recognized as having the same meaning. Also perhaps a jotted sign is the closest I could transcribe two closely produced signs with different meanings. I feel this is close enough for a dictionary entry.

How do you want me to compare two systems? Using my codes and their pictures or using my production description translated into English and their production system translated into English? Will your dictionary entries actually be indexed according to the English descriptions of how a sign is produced, or will it contain encoded / pictured words? I suppose that the real question is: how much ASL / ASL transcription code will a student need to know in order before they can use the ASL dictionary entries? - Tom 07:16, 08 April 2009 (UTC)Reply

An interesting contrast between your system and the one currently used here is that your system essetially is closer to a phonemic transcription system than the loose phonetic transcription system currently used here. I don't know whether you're familiar with the difference between phonemic and phonetic transcription, but an example of an ASL phoneme is the handshape /B/. Most ASL linguists say that it's phonemically irrelevant whether the thumb is extended, flat, or tucked when making the B shape, so the phones , , and are apparently three allophones of the phoneme /B/. That is, a native ASL user typically recognizes two different productions of the sign for BOX as the same sign regardless of the thumbs configuration. In fact, that person typically doesn't even notice where the thumbs are when someone makes the sign.

Dictionary entries may be indexed phonemically, phonetically, or according to whatever tradition a particular language has. Phonemic indexing (Stokoe's approach) works if the phonology of the language is set in stone. Unfortunately, there is surprisingly little agreement among ASL linguists about the phonemic inventory of ASL. For example, consider two signs currently indexed at Index:American Sign Language/H. The thumb extension in signs like (deprecated template usage) H@Chin-PalmBackFingerUp Flatten is irrelevant, so some would consider and as two different allophones of /H/, but then we notice that (deprecated template usage) OpenU@Side-PalmBack Flatten requires . It is misinterpreted as TWELVE if it is made with the shape, so they are not allophones. There are also arguments that phonemic inventory of ASL is variable by region and perhaps even in flux. The lack of wide agreement about the phonemic inventory of ASL suggests that we take the more tenable stance of indexing ASL entries phonetically rather than phonemically. (Of course, it's not quite that simple, since there is a wide range of levels of precision in phonetic transcription, so we start with a minimal list and expand it as needed, making no phonemic claims for now.)

Back to your question, though, “How do you want me to compare two systems?” We could either (a) make a list of your symbols and show which phones from Appendix:Sign language entry names correspond with each of yours, noting questionable comparisons and relevant differences in text, or (b) do that in reverse. The former approach could highlight areas where your system that may not distinguish between two different signs and it would be interesting to see whether we can find signs showing that missing distinction. The latter approach would highlight any contrasts that may be missing from the system in use here. Of course, I'm interested in the latter because that would probably show us where to extend this system. Another benefit of either comparison may be that adopting aspects of your system may reduce the verbosity of the transcription used here.

Another question you ask is, “how much ASL / ASL transcription code will a student need to know in order before they can use the ASL dictionary entries?” The goal answer is, “as little as possible.” Most uses of this dictionary proabably fall into two categories: (a) readers who know an English word and want to know the corresponding ASL sign, and (b) readers who have seen an ASL sign and want to know the meaning. For the former scenario, a reader should visit the entry for the English word (e.g. (deprecated template usage) cute) and see the corresponding ASL signs in the "Translations" table(s). For the latter scenario, a reader would ideally begin at Index:American Sign Language, select the initial shape of the dominant hand, select the location, and click from the list there to the entries that are actually arranged by encoding. Such a user should need no more than basic ASL fingerspelling and basic understanding of the English language. (I don't think that goal has been achieved, but it's a wiki and any improvement is more than welcome.) So, I think the answer to your question, “Will your dictionary entries actually be indexed according to the English descriptions of how a sign is produced, or will it contain encoded / pictured words?” is “yes, and yes, but with a twist.”

I know I haven't addressed all of your points yet, but I have to get back to work. Will follow up later. Cheers! —Rod (A. Smith) 18:44, 8 April 2009 (UTC)Reply

Handshape comparisons

More ASLSJ handshapes

Locations on opposite hand comparisons

Locations on the body comparisons

ASLSJ uses coordinate starting positions and relative movement directions. Few signs are used as exactly as the signs "designer" intended.

Locations in space comparisons

Sign Jotting can optionally use Higher1-3 and Lower1-3 in combination with Left1-3 and Right1-3.

Facing/orientation comparisons

Other features of holds and moves comparisons

Sign Jotting movement styles

Sign Jotting facial expressions (sentence / clause prefixes)

Sign Jotting uses quotes within sentences to locate or transfer to speaking roles. The Left1-3 and Right1-3 are usually first applied to locator signs. The quote is the direction before the quotes specifies to which located object the speaker directing their comments.

>JOHN <CARL <"? Astlvh Dbsgo." >"! Basuio."

John asks Carl, "How are you?" Carl replies, "I'm fine."

- Tom 06:57, 9 April 2009 (UTC)Reply

Very helpful. Thanks, Tom. Following are my takes on some anomalies indicated by your thorough comparison:

• ASLSJ symbol D corresponds to two phones used here: and . I think I've noticed some signers distinguish between (deprecated template usage) 1@FromPalm-FingerUp-1@CenterChesthigh-FingerUp Contact and (deprecated template usage) D@FromPalm-FingerUp-D@CenterChesthigh-FingerUp Contact. An ASLSJ transcription would miss that distinction, right?
• ASLSJ symbol B corresponds to two different phones used here: and . I think the classifier construction FENCE-EXTENDING-FROM-A-TO-B is different from SHORT-WALL-EXTENDING-FROM-A-TO-B only in the spreading of the fingers. If so, there is an important difference between and that ASLSJ wouldn't account for, right?
• Similarly, I think is distinct from in some contexts. E.g., I'm not certain, but I think if the regular were used in the sign for MONEY, I think that would be interpreted as SHORT-SECTION-OF-PIPE-LYING-ON-THE-GROUND. I'm less certain about vs. , but I think that, when showing the quantity of liquid in a container, use of would be misinterpreted. I can't think of any possible confusion between and , so maybe we could drop from the inventory here.
• Ommission of the handshape here was clearly an oversight. Thanks for calling that out. It was included in the index, but somehow missed in the appendix. I added it just now.
• You point out that ASLSJ has the distinction between and (a.k.a. ). I assume that means an shape with the proximal joints (those closest to the palm) of the index and middle fingers bent, but the medial and distal joints fully extended. Is that right?
• The location phones are more difficult for me to categorize. I think there's an important distinction between and , both of which ASLSJ transcribes as bo, but it's hard to think of a good example at the moment, or of any important differences between the other anomalies you made evident in your comparison above.
• As your comparison suggests, and are really similar, and it may be better to consolidate them into one phone for transcription purposes, but they are definitely distinct from , since signs like (deprecated template usage) 1@Chin would be misinterpreted if produced too high.
• An interesting difference in approach is between ASLSJ's <<<, <<, <, >, >>, >>> sequence and the , , , , , sequence used here. The jury is still out on how useful , etc. are for dictionary entry indexing, but the idea is that signs that index real world entities are made with respect to absolute degrees of left and right rather than with reference to the signer's dominant hand. That is, HE-ON-MY-LEFT is indicated by pointing left rather than "to the nondominant side" or "to the dominant side". There are different ways to frame that, though, and it's arguably part of discourse rather than part of lexicology. In any event, angle brackets cannot be used here in entry pagenames for tecnical reasons (due to limitations of the wiki software hosting this project).
• You mention that ASLSJ uses wrist rotation and hand placement to convey the rotation. There's a version of DISMISS that looks like NICE, but with the dominant hand rotated so the palm faces the dominant side. Does ASLSJ have a way to express such a rotation?
• You indicated that ASLSJ doesn't specify any equivalent for , , , , etc. Does that mean that ASLSJ does not distinguish between MONTH and TEMPERATURE (like MONTH, but with a very short movement), or between SLOW and VERY-SLOW (emphatic, slow version of SLOW)?
• For the phones that represent a local movement like , , , etc., it's interesting that ASLSJ uses a sequence of postures. I guess ASLSJ would transcribe (deprecated template usage) FlatC@Forehead Flatten as something like CtcnOtcnCtcnOtcn, but how would it treat something like (deprecated template usage) FlatO@Side-PalmUp Rub or (deprecated template usage) FlatO@InsideChesthigh-FlatO@InsideChesthigh Rub, where the thumb rubs back and the fingertips with several (but an unspecified number of) repetitions? (The system in use here doesn't cleary explain the type of rubbing, e.g. between different finger tips or up and down on a fixed pair of fingers, so there's room for improvement here.)
• It's interesting to note the different approach for movement style taken by the current system here vs. ASLSJ. Here, movement of each hand is specified separately, but ASLSJ takes advantage of the fact that ASL has a limited number of movement styles for lexical words. I don't know how well that works for two-handed classifier constructions, though. Does ASLSJ have a way to express independent hand movement for two-handed classifier constructions (e.g. PERSON-WALKS-AWAY-FROM-MOVING-VEHICLE)?
• It would probably be useful for this project to adopt a version of the ASLSJ facial expression phones. They were left out because it's really hard to enumerate them, and they seem more like the role intonation has in English than the role tone has in a tonal language. Still, they're obviously required in certain contexts, so I think we should do a better job of represent them, as you have done.

As for actual encoding symbols, do you propose that the phone names used in this project be abbreviated to something more like ASLSJ? My impression is that the current system strikes a nice balance between brevity and transparency. I'm not certain how important that transparency is, but it's nice if readers can understand a transcription with minimal training. Too bad SignWriting isn't ready. —Rod (A. Smith) 19:25, 9 April 2009 (UTC)Reply

Truly, ASLSJ has traded off accuracy to gain efficiency. All the points you noted are correct that ASLSJ does not differentiate between the technical implications of meaning. Instead I needed something that could encode an entire sign JUST ENOUGH to be distinguishable from MOST other signs. Therefore it can be used for creating mostly unique dictionary entries. At the end of the day, a dictionary is used to find the words to which you know most of its parts. The fact that it uses keyboard letters makes it easy to specify which parts you know.

But it cannot teach sign language. The dictionary definition of a sign language word would be a sign language sentence in a videos. My problem lies in finding the videos that I want without already knowing the English word. This will likely remain a problem, but the dictionary could lead me to at least a group of likely English words.

I saw someone with their C hand in front of their forehead. The hand moved out and in repeatedly, each time closing to an O hand. ASLSJ would translate this in CsunoiOs.

The construction of ASLSJ has specific encoding quirks. Understanding how to reproduce this code manually would take something like a community college course. (I doubt that there would be many takers at this point.) What I'm sure of is that the more sign details that you throw into the dictionary index the less likely someone will be able to find the entry they want. Bullet points instead of paragraphs.

At the same time, learning my code, your code, or anyone else's "alphabet-convenient" code is not going to happen for most people who have an interest in sign language. They just want to learn the language. As you no doubt agree, Sign Writing is the only computable sign language representation that most people could pick up quickly.

Thus, (at the risk of sounding conceited,) what I would recommend is to take my code principles and assign Sign Writing pictographs for each. The lookup would be visual enough to not even need to know English and it would be general enough to isolate a group of similar signs in a few steps. My code would not be seen except by the computer programs we use to create the slightly-inaccurate Sign Writing indicies. Once the entry is located, you can have it translated from a video to Sign Writing, PSE, and English. - Tom 08:03, 10 April 2009 (UTC)Reply

You say, "The dictionary definition of a sign language word would be a sign language sentence in a videos." That is perfectly accurate for a one-language dictionary, but when the dictionary reader does not know the subject language well enough, such definitions are difficult to use. So, the Wiktionary projects implement an expanded version of that idea. We separate the subject language (the language whose word is being documented) from the meta-language (the language used to document the word). An English speaker can thus use the English Wiktionary (en.wiktionary.org) to find information written in English about an English word, a German word, or an ASL sign. Similarly, a French speaker can use the French Wiktionary (fr.wiktionary.org) to look up the same three words and get meaningful results in his or her language. Today, there is no ASL Wiktionary, but if there were, it would be at http://ase.wiktionary.org/ . It would include entries for English words, German words, and ASL signs. The definitions, etymologies, and other lexical facts about those words would be presented in ASL videos. (There is yet another project that tries to minimize the redundancy that might be bothering you right now, but that project isn't yet able to handle the lexical facts that we want to document.)

In the current system, if a reader sees someone with a C-like hand in front of their forehead, he or she could look up the sign by starting at Index:American Sign Language/C. That index notes the similar shape “Flat C” as having “the fingers and thumb partially bent at the base but otherwise straight”, so the reader would hopefully click through to Index:American Sign Language/FlatC, go to the section “FlatC” in front of or near the head, see the entry (deprecated template usage) FlatC@Forehead Flatten FlatO@Forehead, click through to the entry, and see from the (currently missing) images or videos that the entry is for the sought sign. The entry name describes the sign in an obvious way to me, but that's probably because I've been working with the system for awhile now.

Indexing is a critical component to this project, so I rather like that your recommendation promises to simplify our sign language indices. Let's try a proof of concept. Assign a pictograph to each of your symbols (of course, deaf-blind readers will still need us to retain English phone names as just alt text or whatever) and rearrange one or two of the pages from Index:American Sign Language along the lines you suggest. If the end result seems easy to use from a reader's point of view, your code could then start to maintain the indices automatically. We should also invite User:Msh210 to weigh in on this, too, as he helped introduce ASL indexing here. —Rod (A. Smith) 16:52, 10 April 2009 (UTC)Reply

They look great! —Rod (A. Smith) 05:03, 21 April 2009 (UTC)Reply

I've got to add the images to the entry pages before I can update the index. I'm starting with all the Open-B's. Then I'll add a section to the Open-B index page and see how that works.

Please be mindful that my code approximates some handshape rotations. For example, "late" would end up as an open-B palm-in fingers-down. ASLSJ uses bent-B palm-down, which from a production standpoint would look very similar. The index section I make will be based off these results. -Tom 05:17, 21 April 2009 (UTC)Reply

I've been mucking around a bit lately in index pages, and still don't know how best to organize them, but I am sure that an index doesn't seem to benefit from being exact. I expect ASLSJ approximation will help. —Rod (A. Smith) 05:28, 21 April 2009 (UTC)Reply

Let me know if there's any way I can help you with your project. —Rod (A. Smith) 04:12, 22 April 2009 (UTC)Reply

I have started the layout for http://en.wiktionary.orghttps://dictious.com/en/Index:American_Sign_Language. I realize that there should be separate pages for groupings, but I don't know the best way to do that. I'll be running through the 124 signs I found in the links, adding my flavor of Sign Writing to the entires, and then filling in the index. Any input is welcome. — This unsigned comment was added by Positivesigner (talk • contribs).

Neat. When you say you don't know how best to group signs into separate pages, do you mean you're not sure which signs to put on which pages or do you mean you're not sure how to code up something to group signs in the way you envision? If you mean the former, I'd lean toward grouping ASL signs according to this hierarchy:

1. Initial handshape of the dominant hand. This top-level grouping could be according to precise phonetic shapes or according to rough ASLSJ handshapes (e.g., a single page for 1, D, and G, and another single page for B, flat B, open B, and bent B, etc.). Eventually, we'd like to align with whatever logical groupings SignWriting might use, but I don't know enough about SignWriting to suggest what those groupings might be.
2. Initial location of the dominant hand. This second-level grouping should probably break signs into groups according to the rough, major signing areas used rather than precise phonetic ones. For example, signs made with the initial dominant handshape B (the first grouping) located anywhere on or near the top of the head would all go into one second-level group, while those located on or near the nondominant hand would all go into another group. I'm not sure exactly where to draw the major area dividing lines.

Eventually, the second-level groupings will be large enough to justify their own dedicated index pages, but there aren't enough entries yet to require separate pages per major location. As for organizing those second-level groupings, I suppose orientation or movement type might be used for sort order. Not sure.

I believe it is critical for the index pages to show an English gloss for each sign. Assuming your code can use regular expressions, you should be able to apply something like this Regex to the entry content to find a suitable gloss: \{\{ase-(adj|adv|intj|noun|phrase|pronoun|verb)*gloss\s*=\s*(+) (yes, that regex could be tighter, but I was going for legibility). I think the second capture group should return the first gloss for the sign. Does that make sense? —Rod (A. Smith) 20:26, 22 April 2009 (UTC)Reply

Oh, and I'm not at all saying that you should remove orientation from your sorting. I see now that it's used. Use whatever organization makes sense to you for now. —Rod (A. Smith) 20:29, 22 April 2009 (UTC)Reply

"Eventually, the second-level groupings will be large enough to justify their own dedicated index pages..." Making the second level pages and their links was actually to what I was referring. This is only the second time I've tried using a Wiki. I feel very confident in how I've grouped the signs as the most recognizable way to find them. I will continue adding to this one index page as time permits me. You can e-mail me at my e-mail address listed on ASLSJ.com for a less formal-feeling discussion. Posting every comment on the internet makes me comment less frequently. — This unsigned comment was added by Positivesigner (talk • contribs).

Replied via e-mail. —Rod (A. Smith) 16:36, 23 April 2009 (UTC)Reply

FYI, I posted an invitation to WT:BP for feedback on the index reorganization effort. I invited them to discuss the topic at Index talk:American Sign Language#Organization. —Rod (A. Smith) 22:06, 27 April 2009 (UTC)Reply

To change the template that appears at the top of the reorganized index pages, edit Template:index/American Sign Language ASLSJ. —Rod (A. Smith) 16:38, 28 April 2009 (UTC)Reply

Thanks for making those. Here is a description of which handshapes are assigned to which group.

• B / C group has first, middle, ring, and pinky fingers away from the palm, possibly one finger bent, and no fingers touching the thumb; B / 4 / open-B / flat-B / 5 have fingers straight, C / claw-5 / flat-C / bent-B / closed-B (thumb-under) have no fingers straight
• F / O group has first, middle, ring, and pinky fingers away from the palm and one finger bent or touching the thumb; F / open-F / 8 / open-8 / 7 have fingers straight, O / E have no fingers straight
• W / M group has first, middle, and ring fingers away from the palm and possibly one finger bent; W / 6 / open-W have fingers straight, M / open-M have no fingers straight
• V / N group has first and middle fingers away from the palm and possibly one finger bent; V / 2 / open-V / 3 / U / R / open-R / K (P) have fingers straight, N / bent-U / bent-V / double-X / open-N / closed-N have no fingers straight
• D / T group has first finger away from the palm, possibly bent, and possibly extending the pinky finger; D / 1 / Corna (1 and I) / L / ILY (L and Y) have fingers straight, X / open-X / bent-D / bent-1 / T / open-T / G / baby-C / baby-O have no fingers straight
• I / S group has no fingers away from the palm and possibly extending the pinky finger; I / Y have fingers straight, S / A / open-A have no fingers straight

The reason for these groups is that just about any handshape choice within the group will keep the sign word recognizable within a context. My CODA teacher always uses D for the number 1. Straight fingers on any handshape may end up being bent anyway (more comfortable as arthritis sets in) making absolute distinctions a matter of dictionary work and not an everyday practicality. They need to be grouped this way in order to find the "correct" handshape that we list. With regional variances, the same sign word may be made under multiple, different handshape combinations. -Tom 10:07, 29 April 2009 (UTC)Reply

Thanks for the explanation of the groupings, Tom. I've added those shape group descriptions to the relevant index pages.

I cannot find much existing research that groups signs like you do. I suppose the next step is to test the effectiveness of the new organization. It might be overkill, but one way would be to show readers images or videos of signs and invite them to try to locate the entries using the index. For example, we could create a list like the one at User:Rodasmith/BC student photo entries, and the reader to find a few such signs using Index:American Sign Language. —Rod (A. Smith) 17:43, 4 May 2009 (UTC)Reply

I just wanted to mention that the handshapes used in ASLSJ do not have a fraction of the variability as SignWriting. The "fingers-up" handshapes are palm-out (black-filled) and palm-in (white-filled) orientations. The "fingers-down" handshapes are supposed to mean "fingers-forward" palm-down (black-filled), palm-up (white-filled), or palm-side (half-white / half-black filled). -Tom 09:35, 5 May 2009 (UTC)~Reply

Could you have a look at Wiktionary:Votes/bt-2009-10/User:Di gama bot for bot status please?​—msh210℠ 18:27, 19 October 2009 (UTC)Reply

Hi, Positivesigner.

I'm not sure I get the point of the "production grids", e.g. at http://en.wiktionary.orghttps://en.wiktionary.org/w/index.php?title=S@InsideChesthigh-PalmUp-S@InsideChesthigh-PalmUp_5@InsideChesthigh-PalmUp-5@InsideChesthigh-PalmUp&diff=8092003&oldid=6463624 . I think the "G" in "GUH" probably means "greater", i.e. "dominant", and the "L" in "LUH" means "Lesser", but without any guide, it's tough to make sense of it. Are the "production grids" meant to be read by people or by software? If they're meant to be read by people, are they supposed to clarify something that's missing from the descriptive text above it?

GUH and LUH were (unpublished) acronyms I made for Greater-Used Hand and Lesser-Used Hand. I was trying to create a nomenclature that would not need to specify Right-Hand or Left-Hand because left-handed people sign the same way. Dominant and Non-dominant are linguistically accurate, but not really easy to understand for passers-by. I am now thinking of using Near-hand and Far-hand / Near-side and Far-side because these is shorter than the spelled out GUH / LUH but still convey the signing-space partiality.

Any edits I made 1 1/2 years ago were to give examples. As you have seen the examples, you may remove them from future versions. I was not trying to set a precedent. The same goes for whatever Sign Writing I added to some of the pages or changes to any of the indexes. They were manually throw together without templates.

— This unsigned comment was added by Positivesigner (talk • contribs).

So, in your estimation, the terms "near" and "far" are more immediately obvious to people unfamiliar with ASL linguistics than the terms "dominant" and "non-dominant". Interesting. I'm surprised, but have only my own opinion to go by. We should try to be consistent in our terminology, though. To decide whether Wiktionary should adopt the "greater"/"lesser" or "near"/"far" terminology, we could start a new conversation in the Beer Parlour. —Rod (A. Smith) 15:37, 21 April 2011 (UTC)Reply

The reason I ask is that, if your grids turn out to add some helpful detail for readers, we could improve them, e.g. by documenting them somewhere and adding links from them to a page (or pages) that explain them. —Rod (A. Smith) 19:47, 18 April 2011 (UTC)Reply

For comparison, I added a "hold-move" chart for MANY. Click the hold-move chart's "show" button to see the details. I think it's a bit overkill, but it's based on the same work by Liddel and Johnson. One nice feature of the hold-move charts is that they default to collapsed view so they don't take up too much of the page. (We could add such a collapse-show feature to your "production grids" if we keep them.) Another feature is that they unambiguously describe all of the phonetic details of each sign in a way that's computer-readable. I just don't know yet that there's any real advantage to the hold-move charts, so I haven't put them in all the ASL entries. Like the long entry name system, we may abandon hold-move charts when SignWriting comes into its own. —Rod (A. Smith) 21:39, 18 April 2011 (UTC)Reply

How can it be determined whether any feature of the ASE entries "adds some helpful detail for readers." Who are the expected readers? "Posture the nondominant hand in the “S” handshape about half arm's length in front of the nondominant side of the chest, nondominant palm facing up" may be 100% accurate, but it sure excludes the common folk.

To determine whether our entry features are useful, we do sometimes get feedback from users, but mostly we just use our collective judgement solicited through the Beer Parlour. If your production grids seem to you to contain useful information that's otherwise missing, or if they have some other benefits, let's discuss so we can either promote your production grids to a standard here or incorporate the benefits of those grids into an existing standard here. I personally just don't get them because they seem to omit important phonetic details (i.e. readers cannot use the grid to learn how to make a sign because they don't seem to say where to position the hands) and they don't seem any more clear to me than the production narratives. You say the narrative "sure excludes common folk." is that due to word choice or what? I would love to improve the narratives so that they no longer confuse casual readers.. —Rod (A. Smith) 15:37, 21 April 2011 (UTC)Reply

The "common folk" I know that are interested in ASL would not learn ANY narrative production code, especially a terse one. I cannot recommend a way to improve the existing hold-move charts' phrasing because: 1) I'm focused on domestic readers, 2) I have not truly studied how to produce the hold-move chart or the title abbreviation and 3) I cannot visualize the audience for whom they are intended.

I believe that "near" and "far" are not readily obvious. I believe the IPA is not readily obvious even to most native English speakers. If they have started teaching the IPA to children in school, I have not heard. That is why english dictionaries have a pronunciation guide that is simpler than the IPA. The hold-move charts are as complicated as the IPA in order to serve the same purpose: tell how a word is said in a particular dialect so that the dialects can be compared.

American English dictionaries' pronunciation guides are specific to American English. Therefore Britons, Irish, etc. would feel our IPA pronunciation is incorrect; and vice-versa. Wiktionary apparently tries to appeal to an international audience, and therefore champions the IPA instead of the generalized national versions. With that view in mind, the hold-move charts are better aimed at Wiktionary's international audience of signers and multiple signed languages.

I cannot imagine any casual readers of ASL Dictionary entries trying to learn a description code instead of looking at pictures or, preferably, a video. There is no other way to readily understand what you are talking about. I remember being 12 years old and checking out an ASL dictionary with line drawings from the library. I wanted to casually read about ASL words. I remember seeing an entry for WANT. The drawing made sense to me with the faint/dotted starting hands and the bold ending hands. I could copy it with my hands. When I produced the word myself, I could clearly relate the English word with the ASL emotion.

Although all the other pictures in the book were equally intelligible, I could not relate the English word's meaning to the production. There was no explanation of the reasoning behind the shapes and actions. Twenty years went by before I took an ASL class in which the teacher related each word's etymology, those pieces of the sign that intimated the real-life pictures from which they were drawn. Only then did ASL actually click for me. My goal is to make it a little easier for others get started down that road. Making a more concise "pronunciation guide" for ASL was in line with that goal, which is why I made my ASLSJ code. The "near" / "far" descriptors are an expanded form of my code's symbols.

I realize that my goal does not match Wiktionary's goal. The reason I want to add more ASL entries is to use them as a reference on my own site. I know quite a few ASL words and can quickly transfer sets of still images from video for upload to Wiktionary, and create new entries from these. I would like to be able to add ASL etymological explanations to Wiktionary as well.

Here are the benefits I see to having a production description. 1) Someone sees a video but is not quite sure about some parameter like the handshape, touch direction or placement on the body and wants to quickly look it up. This is how I use the printed dictionary if I have seen a word in print and want to know how it is spoken. 2) Someone is comparing / contrasting two different word's production descriptions or two productions for the same word.

When I want #1, I use a general description code that I can grasp quickly. But I currently never see ASL in print except for my own writing, and so it is not applicable to Wiktionary. The current hold-move descriptions are designed for #2, and does an adequate job of it. I just cannot see that the casual ASL Wiktionary browsers would try to learn any description code. Therefore I figure that these descriptions must be for some specific audience, but not the general domestic audience that I usually envision. That doesn't make it bad; just less likely to be used. -- Positivesigner 05:28, 23 April 2011 (UTC)Reply

Hopefully SignWriting will be in Unicode soon and a widely available font will include it so we can start using as concise descriptions of signs per your #1 above. Have you seen the sign language toolbox at the bottom of the page when you're editing in Wiktionary? If not, click "Edit" here and then choose "Sign languages" at the bottom of the screen (below "Save page"). That may help you while we're waiting for SignWriting to get into Unicode. This Wiktionary serves both casual and academic users, so videos, links to YouTube, and any other suggestions you have are welcome. It takes time to build a crowd-sourced project like this, so any entries you create or expand would be appreciated. If you don't know the perfect name for a sign per WT:ASGN, don't let that hold you back from contributing.  :-) —Rod (A. Smith) 06:41, 23 April 2011 (UTC)Reply