What you describe is only true in morpheme-based approaches to morphology (what are also referred to as item-arrangement approaches). There are also item-process approaches and (increasingly in the last 50 years) word-paradigm approaches (see Blevins, 2016). Stump 2001 has a different taxonomy but also presents alternatives to a morpheme-based approach focused more on paradigms.
On Wed, Oct 25, 2023 at 2:32 PM Hugh Paterson III via Corpora < corpora@list.elra.info> wrote:
As I have followed this conversation (and started it) I have often wondered why the term lemma is used at all. In my understanding of morphology there are roots and stems. My understanding is that roots are single morphemes whereas stems may be composed of more than one morpheme. Then additionally one has affixes (with the generally accepted subcategories being: prefixes, suffixes, infixes, circumfixes). In some theories of word construction there are 'bases'. Also some people accept allomorphs, which may have some morphophonemic impacts on the stem/root. I'm not at all clear what the theoretical implications of a lemma infer. If someone can point me to the literature on the difference between a lemma and a root that would be fantastic. I had supposed that lemma was the computational term for a root. But maybe this is not the case? Given that we have lemmas in various ontologies such as is used by LiLa and NIF do these formal categories also have a formal definition or test which can be applied cross-linguistically to determine if the content indicated to be a "lemma" actually fits the categorical definition? In information science we call these sorts of requirements content standards which apply at the semantic level (rather than some constraint on the syntax or composition of the string).
Kind regards, Hugh
On Wed, Oct 25, 2023 at 7:25 AM Ada Wan via Corpora < corpora@list.elra.info> wrote:
Dear Orhan
Thanks for your reply.
You don't have to do both ('wean ourselves of the concept of a lemma and at the same time think of "a verb that can be conjugated"') at the same time. You can do one after the other, your choice. And I also didn't mention anything about "lemma" for the thought experiment. I would like one to understand (linguistic) morphology first, above all, the idea behind such. Perhaps one's addiction to lemma/lemmatization has to do with a lack of understanding?
Re "lemma": in the context of our discourse, the term and the use of "lemma" only has a few decades of history. One of the goals of education is to train students to become mentally agile. It is fine to take a concept as an assumption and develop theories therefrom (assuming it's ethical and appropriate to do so). But one needs to be able to not use it whenever timing is appropriate, to not become addicted to it, to abandon it whenever there is a better, less biased alternative that can supplant it... etc.. A good researcher also ought to learn to develop a sense for such timing and context, including but not limited to when to self-correct, when to let go of a theory/hypothesis.
Re "not allowed to talk about things": in general, given the freedom of expression, one can talk about anything one likes --- but if this were to relate to education/research/teaching, I'd ask that one thinks twice about what one claims. There has been some miseducation and/or bad theories/hypotheses in the "language space" (above all, in structural linguistics and the computational variant thereof). This is the main point I tried to address with my "objection" to "morphological endeavors".
Re using the term "lemma" to refer to things that can be "conjugated/inflected": because "lemma" does not refer to things that can be "conjugated/inflected".
Hope that helps. Looking forward to your answers to my questions.
Thanks and best Ada
On Wed, Oct 18, 2023 at 9:03 PM Bilgin, Orhan (Postgraduate Researcher) < o.bilgin@lancaster.ac.uk> wrote:
Hi Ada,
Thank you for your reply.
I don't think it is possible to follow your advice to wean ourselves of the concept of a lemma and at the same time think of "a verb that can be conjugated", because that is precisely an example of what I would call a lemma.
I never claimed that anything exists beyond the reality of my mind. I only asked why I am not allowed to talk about things that can be conjugated / inflected etc. and to use the word "lemma" to refer to those things. You haven't answered that question.
Best,
Orhan
On 18 Oct 2023 17:49, Ada Wan adawan919@gmail.com wrote:
[To those who do not have shared interests on issues that pertain to Corpora-List matters, such as data/corpora and their handling which includes but is not limited to linguistic/NLP theories/methods (and the validity thereof): please disregard.]
Dear Orhan
Thanks for your interests in this discussion. I think it is high time that our community comes to a critical (re-)examination of (linguistic) morphology (and to address issues concerning reinterpretation and transition).
First of all, allow me to put my traditional grammarian hat on to get to your question more directly. You brought up an example of a morphological paradigm. Now, as linguists or language professionals, we know that language is (re-)productive in nature. So, if you don't mind, we can do a thought experiment and go through this dialectically (pls note that I only check my emails about once a day on weekdays, however).
- Let's think of a verb that does not yet exist (in any particular
language(s) that you can think of or that you are used to). Would you mind conjugating it for me? How many patterns would you have? And what would the forms be like? 2. Where did you get the patterns/paradigm from? If you were able to come up with a "full paradigm" (whatever that should refer to (?) --- but let's suppose, you have 6 forms (as per some textbook paradigms from some "Indo-European languages" --- 1st/2nd/3rd person in sg/pl), you surely haven't seen any of these forms combined with the verb before, have you? So where is your evidence that these forms exist in reality beyond that of your mind? And if such "perfect/ideal paradigm" exists only in your mind (and minds of some of your friends as well), how do you justify that morphological paradigma (the form/"structure"/pattern) are a necessary or intrinsic part of language (may these be of any particular language (which "one"?) or or language in general)? Wouldn't morphology as well as the perpetual construction and reconstruction of morphological patterns be a self-fulfilling prophecy only? And how often do we impose our conceptual/perceptual habits/categories upon whatever "new" that we encounter? 3. If, however, you were not able to construct a "full paradigm" or any part thereof at all, or you claim you were not able to think of a hypothetical verb either, because to you morphology is solely based on what has been written and analyzed beforehand/historically, then what is there to claim about morphological analyses? Not only does such practice not generalize, but it would also just apply to calcified segments analyzed/interpreted in a certain way as part of philological pursuits in the past. One should bear in mind that philological methods can progress and update as well.
There are no limits as to how one can *use* (or some might even claim *define* here) "language", including how various modalities can combine/fuse with each other. Meaning has no fixed boundaries. When it comes to language or meaning, there is no "completeness" to "speak of" or to serve as basis of any science/study. And there are no fixed demarcations between any "particular languages" either.
Other perspectives on (the shortcomings of) morphology and "words" can be found on my rebuttal page here: https://openreview.net/forum?id=-llS6TiOew. Please also read the references cited therein.
I look forward to your reply, comments/remarks, or questions. (Actually, the floor can also be opened to anyone who would like to join.)
Thank you and best Ada
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
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