Hi Anil
Thanks for your comments. (And thanks for reading my work.)
Yeah, there is a lot that one has to pay attention to when it comes to what "textual computing" entails (and to which extent it "exists"). Beyond "grammar" definitely. But experienced CL folks should know that. (Is this you btw: https://scholar.google.com/citations?user=QKnpUbgAAAAJ? If not, do you have a webpage for your work? Nice to e-meet you either way!)
Re "I know first hand the problems in doing NLP for low resource languages which are related to text encodings": which specific languages/varieties are you referring to here? If the issue lies in the script not having been encoded, one can contact SEI about it ( https://linguistics.berkeley.edu/sei/)? I'm always interested in knowing what hasn't been encoded. Are the scripts on this list ( https://linguistics.berkeley.edu/sei/scripts-not-encoded.html)?
Re the unpublished paper (on a computational typology of writing systems?): when and to where (as in, which venues/publications) did you submit it? I remember one of my first term papers from the 90s being on the phonological system of written Cantonese (or sth like that --- don't remember my wild days), the prof told me it wasn't "exactly linguistics"...
Re "on building an encoding converter that will work for all 'encodings' used for Indian languages": this sounds interesting!
Re "I too wish there was a good comprehensive history text encodings, including non-standard ad-hoc encodings": what do you mean by that --- history of text encodings or historical text encodings? After my discoveries from recent years, when my "mental model" towards what's been practiced in the language space (esp. in CL/NLP) finally *completely *shifted, I had wanted to host (or co-host) a tutorial on character encoding for those who might be under-informed on the matter (including but not limited to the "grammaroholics" (esp. the CL/NLP practitioners who seem to be stuck doing grammar, even in the context of computing) --- there are so many of them! :) )
Re "word level language identification": I don't do "words" anymore. In that 2016 TBLID paper of mine, I (regrettably) was still going with the flow in under-reporting on tokenization procedures (like what many "cool" ML papers did). But "words" do certainly shape the results! I'm really forward to everyone working with full-vocabulary, pure character or byte formats (depending on the task), while being 100% aware of statistics. Things can be much more transparent and easily replicable/reproducible that way anyway.
Re "We have to be tolerant of what you call bad research for various unavoidable reasons. Research is not what it used to be": No, I think one should just call out bad research and stop doing it. I wouldn't want students to burn their midnight oil working hard for nothing. Bad research warps also expectations and standards, in other sectors as well (education, healthcare, commerce... etc.). Science, as in the pursuit of truth and clarity, is and should be the number 1 priority of any decent research. (In my opinion, market research or research for marketing purposes should be all consolidated into one track/venue if they lack scientific quality.) I agree research is not what it used to be --- but in the sense that the quality is much worse in general, much hacking around with minor, incremental improvements. Like in the case of "textual computing", people are "grammar"-hacking.
Re *better ... gender representation": hhmm... I'm not so sure about that.
Re "About grammar, I have come to think of it as a kind of language model for describing some linguistic phenomenon": nah, grammar not necessary.
Re grammaroholic reviewers: yeah, there are tons of those in the CL/NLP space. I think many of them are only willing and/or able to critique on grammar. Explicit is that it shows that they don't want to check one's math and code --- besides, when most work on "words" anyway, there is a limit to how things are replicable/reproducible, esp. if on a different dataset. The implicit bit, however, is that I think there is some latent intent to introduce/reinforce the influence of "grammar" into the computing space. That, I do not agree with at all.
Re "magic": yes, once one gets over the hype, it's just work.
Re "I have no experience of field work at all and that I regret, but it is partly because I am not a social creature": one can be doing implicit and unofficial "fieldwork" everyday if one pays attention to how language is used.
Best Ada
On Sat, Aug 5, 2023 at 8:51 AM Anil Singh anil.phdcl@gmail.com wrote:
I forgot the main reason for writing the last email. Most importantly, I share your view that orthography is underrepresented in NLP/CL. I had once tried to build a computational typology of writing systems. The paper was not published, but I still believe that is something worth doing. Perhaps one day I will complete that work.
Also, I am conscious that, technically, I used the term category mistake in a wrong way, but I hope I was understood correctly.
On Sat, Aug 5, 2023 at 12:47 AM Hesham Haroon heshamharoon19@gmail.com wrote:
Hi Ada and Anil,
I'm enjoying reading your discussion. It's been very informative and thought-provoking. Thanks for sharing your insights!
Best, Hesham
On Fri, Aug 4, 2023, 8:51 PM Anil Singh via Corpora < corpora@list.elra.info> wrote:
I have been enjoying the discussion. I hope it will continue. I have learnt some new things. I was also confused about the tensor thing, although not in the same way.
I hope I am not among one of the scare quoted NLP practitioners, because that's exactly what I like to call myself. I certainly don't think I am qualified to work on language just because I can speak one.
I am currently reading your thesis and trying to digest it.
I also glanced through the syllabus you are preparing. I share your interest in text encodings. among other things. I can't resist talking about text encodings, whether I am teaching NLP or Computer Programming, because I know first hand the problems in doing NLP for low resource languages which are related to text encodings.
If you can actually teach that syllabus, I envy you as I am unable to get people interested in the very basics of language/linguistics.
About the importance of granularities, I had, in my (very badly written) PhD thesis, explicitly talked about NLP problem formulation in terms of granularities. In my second research paper, I had used byte n-grams for language identification. I use byte n-grams whenever I can. Actually, I used it for language-encoding pair identification, as there are so many non-standard 'encodings' which were used and perhaps are still used for South Asian languages. My very first -- unsuccessful or you may say unfinished -- attempt at doing some kind of NLP even before knowing that a field called NLP or CL existed, was on building an encoding converter that will work for all 'encodings' used for Indian languages. I too wish there was a good comprehensive history text encodings, including non-standard ad-hoc encodings.
I also share your interest in word level language identification. In 2007 I had published one of the earliest papers on what I called language identification in a multilingual document, where I had tried word level language identification, and what is now called language identification for code switched data.
About gender, I had actually made a kind of category assumption. I didn't pay attention to the name, which you share with no less than Ada Byron.
We have to be tolerant of what you call bad research for various unavoidable reasons. Research is not what it used to be. At least that's my opinion. Still, in some ways it is better, perhaps like in the case of gender representation.
About grammar, I have come to think of it as a kind of language model for describing some linguistic phenomenon. I once received a review in which the reviewer mentioned some grammatical mistakes and wrote that you don't have to just see how the sentence/phrase sounds, you have to explicitly check the grammar according to the rules. Thank you very much, but I learnt English without paying any explicit attention to grammar. I am pretty sure I didn't learn much from explicit teaching of grammar, whether of English, or of Sanskrit, or of French.That doesn't necessarily mean I don't believe in grammar, but I guess I am moving towards the language games view of language.
As to language being magical, well, that depends on what you mean by magical. To me, it seems it is magical in the same sense as life itself is magical. Nothing more, nothing less. Even computer programming I have been known to call magical in a certain sense.
I also completely agree that we can only hope that we are communicating as we intended, but we rarely, if ever, actually attain that goal.
I can't match your background, but I did have -- what can be called -- four rounds of graduate training in different disciplines. I am still trying to learn new things about language. However, I have no experience of field work at all and that I regret, but it is partly because I am not a social creature, or, to be more precise (as if one can be precise with language), I am socially totally incompetent. I wouldn't know how to approach anyone for fieldwork in Linguistics.
On Fri, Aug 4, 2023 at 9:03 PM Ada Wan via Corpora < corpora@list.elra.info> wrote:
@Toms: for completeness' sake: would you mind please sharing your background? Thanks.
On Fri, Aug 4, 2023 at 5:31 PM Ada Wan adawan919@gmail.com wrote:
Thanks x2, Ibrtchx.
On Fri, Aug 4, 2023 at 3:30 AM Albretch Mueller lbrtchx@gmail.com wrote:
On 8/3/23, Toms Bergmanis toms.bergmanis@tilde.lv wrote: ...
I, for one, have benefited from Ada's, as well as other member's suggestions and comments as I hope they have somehow benefited from mine. lbrtchx
Corpora mailing list -- corpora@list.elra.info https://list.elra.info/mailman3/postorius/lists/corpora.list.elra.info/ To unsubscribe send an email to corpora-leave@list.elra.info
--
- Anil
Corpora mailing list -- corpora@list.elra.info https://list.elra.info/mailman3/postorius/lists/corpora.list.elra.info/ To unsubscribe send an email to corpora-leave@list.elra.info
--
- Anil