Implementing a technical "dictionary/glossary" for non-English languages.

I think this has come up before, maybe tangentially on some specific site (EE.CD ?), but I lost track of the posts/comments.

Anyway, I think it would be useful, at least for EE.CD, but I post here on Meta because it could be useful for other communities, especially those focused on science and technology.

My idea is that it would be useful for a community to have a place where to build up a technical dictionary/glossary of the terms found in their field(s) with translation in other languages.

Nowadays English is the lingua franca for science and tech and people searching our site come potentially from anywhere in the world. As a teacher that has been stressing the importance of learning English to his students since 2005, I would find very useful a page (on, say, EE.CD) with a list of all English technical terms and the corresponding translation in Italian, maybe with usage suggestion to avoid false friends and the like. That would be an awesome resource for students and even for professionals wanting to write technical documents.

The availability of such a resource could even promote our site. As far as I know there is no site that offers this service for free (as in free beer). There is Wikipedia, but it's not organized as a dictionary and searching it just for terminology is not so easy. Moreover the coverage is more spotty, since Wikipedia is a generalist site, whereas ours are focused communities where you could ask someone for advice and clarifications.

Therefore I thought to throw in this half-baked "proposal" (more of a brainstorming session, maybe).

Just to get things started I think that such a tool could be implemented in a very basic fashion using the existing facilities. If a community created a new category (let's say "Dictionary"), we could have a post for every language and that post would contain a list of terms with their translations and maybe usage hints. That post would be edited with updated/new definitions as needed.

How would that play out?

As an example off the top of my head, here is a mockup of a post:

English → Italian (Italiano)

• array antenna → antenna a schiera.

• ADC (Analog-to-Digital Converter) → convertitore Analogico-Digitale; convertitore A/D.

  L'acronimo inglese ADC è usato invariato anche in italiano.
  
  The English acronym ADC is used as is also in Italian.

• BJT (Bipolar Junction Transistor) → transistor bipolare a giunzione; transistor bipolare; transistor(e).
  Il termine italiano transistor(e) è usato spesso impropriamente come sinonimo per indicare il transistore bipolare. In realtà il termine transistor(e) è più generale ed include anche altri dispositivi.
  
  The Italian term "transistor(e)" is often used improperly as a synonym for bipolar transistor. Actually the term "transistor(e)" is more general and encompasses also other devices.

• capacitor → condensatore;
  Non usare il termine capacitore. Qualche autore ha cercato in passato di usare una traduzione diretta del termine Inglese moderno "capacitor", ma non ha avuto successo. Il termine tecnico standard corretto in Italiano è condensatore.
  
  Do not use the term capacitore. Some author has tried in the past to use a direct translation of the modern English term "capacitor", but with no success. The correct Italian standard technical term is condensatore

Of course if we had more specific tools it would be more streamlined to add definitions, but devs resources are scarce now, so we could get by with a more "manual" approach. The only problem with this is to devise a format that could be easily transferred into a database in the future.

What do you think?

EDIT

To address a comment by Derek Elkins, expressing a legit objection:

This seems a bit contradictory. If dictionaries "really need coordination and vision", then why would building a dictionary here work? Why would this not be just as bad Wiktionary? Is just because it would be more technically focused (and more focused overall)?

I was trying to reply in a comment, but it turned out it was unwieldy, so probably the answer belongs here.

First of all, I never proposed to create a "real" dictionary, i.e. a work with all the professional linguistics expertise that that would require. In fact I said "dictionary/glossary" in quotes (for want of a better short term).

One of the problems of creating a good, professionally curated dictionary is that linguists are not domain experts, so creating a huge corpus of terms, with linguistic information, cross references and grammatical examples anyone can understand is an incredible feat.

For example, did you know that good "real" dictionaries have their definitions written using a restricted set of words so that anyone knowing just that set can understand any definition? Any definition that cannot be written in a meaningful way in that common set has to be carefully vetted and any extra words are often marked as "special" in the definition itself (so a reader can infer the prerequisites of that definition). That's a feat in itself. OK, computers can help in that area, but dictionaries were written well before computer age, and anyway computer can spot a "non-basic word", but can't automatically substitute that word with something that has the same exact meaning in that exact context. That requires a human (if AI changes that, we will see).

To the point now. Our effort may be more successful than Wiktionary (hopefully) for a bunch of reasons, IMO.

1. Ours would be focused on a specific field and targeted at people already knowing English (at least to a point). Hence we wouldn't need to be concerned to explain grammar structures to English newbies. Moreover we wouldn't need to be constrained by a basic set of words. We would also take for granted that users know the target language of the translation. For example, if you browse the Italian entry you should know both English and Italian, at least to a reasonable level. It isn't meant to be a learners' Dictionary.

2. Ours would be an aid to correct and meaningful technical translation, not an encyclopedia of technical terms. We would take for granted the concepts in the glossary were already clear. We wouldn't have to explain what an array antenna or a transistor is, for example. If a user doesn't understand the concept behind a term, they should learn that first.

3. It would be curated by domain experts that are publicly accountable (through our rep system and the privilege system). Moreover we have a public "moderation" area (community-specific meta), where problems could be brought up (does Wiktionary have this?). The focus here is on "public". Anyone can see the process going on without being logged-in. From a security engineering POV, ideally we have a platform that supports the creation of a web-of-trust among its users that also involves viewers, and this could play a role in curbing bad content.

4. We have a smaller editors community "built-in" in how our system works, so it's less likely that a random guy on the Internet wants to chime in and add just a definition and then leave the site forever. They could leave a comment, though, and if it is relevant hopefully a curator would pick that info up and update the content.

5. Smaller "dictionary" size. We are talking about, what, 1000-2000 words maximum per field/community (disclaimer: I have tech/science-focused communities in mind; for others like Outdoors or Music this maybe is not applicable. IDK)? "Real" dictionaries begin at 10 times that size. The pocket-size dictionaries from reputable publishers (e.g. Collins, Oxford, Cambridge, Duden, just off the top of my head) that you may see around are just an excerpt of a much bigger work. Whereas other smallish "tourist dictionaries", often embedded in tourist guides from less committed publishers are usually crap in my experience (they could let you get by in a day to day situation on holiday, but in extreme cases they could also get you arrested because of some false friend and hasty translation).

6. Focused audience. Active users would be experts, enthusiasts or students in the field, so flagging bad content could be more effective.

posted 9 months ago

CC BY-SA 4.0

9mo ago by trichoplax‭

Lorenzo Donati‭

14 3 63 20