An Empirical Investigation into the Nature and Degree of Nominality – Alex Carr

For an enlarged version, click on the poster or download the file:

Abstracts page

Bios page

Problems/Difficulties:

1.Are there any specific features that you think would be interesting to explore in my distributional semantic model, apart from ‘5 collocates each side of the UEN’?

2.As my visit at KU Leuven was cut short, I did not get to fully learn about distributional semantics. Do you know any relevant literature, or online tutorials which focus on setting up a distributional semantic model?

Comments

  • Lucy Chrispin

    Hi Alex, I have one idea in response to question 2. ESRC offer additional funding and one thing is the overseas institutional visit – http://walesdtp.ac.uk/funding/overseas-institutional-visits/. Obviously this is very dependent on when you need training and the pandemic situation, but you might find training at a particular institution or arrange with a particular researcher. I’ll keep an eye out for other ideas too.

  • Alex Carr

    Hi Lucy,

    Cheers for the comment. I had thought of potentially applying for funding for an overseas institutional visit when the time was appropriate. So I’ll definitely consider that, pending the end of the pandemic of course! Thanks, again!

  • Gerard O'Grady

    Hi Alex, interesting work. One word of caution is that an SFL experiential analysis and work by people such as Geeraerts are not so easy to bring together. They come from different places and have very different aims. One major weakness in Geeraerts work is his lack of interest in temporality. More generally I sense some tension between your aims and your data set. You as I understand it are investigating nominality in English which is a very different thing from an investigation of nominality. The work of west coast functionalist such as Chafe, and Givon may be of relevance here. Finally I wonder whether you need a diachronic aspect to your work. As you mentioned that you were due to go to KU Leuven a scholar whose work might be of use to you would be Jean-Christophe Verstraete – he doesn’t work with English but his ideas on nominality might be of interest.

  • Alex Carr

    Hi Gerard,

    Thank you for commenting! That’s interesting about the clash between Geeraerts work and SFL. I have generally used Geeraerts work in my thesis so far to establish the multidimensionality of meaning and how a semasiological perspective of word meaning can inform how nominals are able to express the meanings they do. Would the clash be that this perspective comes from a more cognitive view of language, whereas SFL places emphases on language as a system?

    I can also see what you mean about this tension. I will make my focus on ‘nominality in English’ more explicit, and give some of the work of Chafe and Givon a read!

    A diachronic aspect would be interesting. I will discuss it with Lise at our next meeting. Sadly, I’ve already been to Leuven (not to say I wouldn’t go again), so that was a bit of a missed opportunity there! I’ll definitely look into his work though, and maybe drop him an email.

    Thank you once again.

  • Kate Kavanagh

    Hi Alex, thanks for the poster! There’s quite a lot of unfamiliar terminology in there for me, but the intro and info on semasiology and lexical aspect are really interesting. I definitely have a better sense of what you’re researching now! I am a little bit confused by the idea of ‘5 collocates either side of the UEN’… it sounds like you are talking about the co-text that appears before and after each occurrence of the UEN (5 lexical tokens either side)? In corpus linguistics, where I’m on slightly more familiar ground, the term ‘collocate’ is used differently, but I think both approaches could potentially be used to feed into your 10000 “features”. Maybe colligation could also be interesting?

  • Alex Carr

    Hey Kate,

    Thank you for giving some time to (and persisting with haha) my poster, as, like you say, it is quite terminology heavy! Glad you found it interesting! Yes you’re right, that’s actually a mistake on my part. So 5 lexical tokens each side of the underived event nominal will constitute one feature, of which there will be 10000. Then each of those lexical tokens will be imputed into a model that will be able to tell me the semantic distance between the underived event nominals, and also whether any underived event nominals cluster together. The idea of this method takes the notion that ‘you shall know a word by the company it keeps’. Thanks again for the comment!

  • Katy Jones

    An fantastic poster Alex and you’ve done an excellent job of explaining some complex terms and concepts to your audience. I might add Peter Fries and John Du Bois to Gerard’s recommendations of people to read.

  • Ellie Bristow

    Hi Alex, fab poster! Really interesting to read more about your work (and I get it more now!!!). I will reply to your message about this (writing this message has just made me realise I haven’t, sorry!), but was just wondering if there are some words that are difficult to determine etymologically or the roots of the word are ambiguous, will this affect your approach/ability to analyse that specific word?

  • Alex Carr

    Hi Katy,

    Thank you! Trying to make all the content fit onto a poster was certainly a task! And thank you, I’ll add them to the list!

  • Alex Carr

    Hey Ellie,

    I’m glad the poster has made it a bit clearer. Always funny trying to explain the topic to people in person haha.

    That’s okay! These difficulties have come up. I’m currently exploring a small data set for a paper I’m looking at writing, so I’ve luckily had the opportunity to investigate these issues on that data, and realise certain problems before undertaking it in my thesis. As you know, it has sometimes been tricky to analyse words that have an unknown/uncertain etymology. This is still something I am looking into, but have thought about taking into account their morphological complexity as a indicator of their etymology. But I’m still working on it. Another problem which has arisen has been words whose morphological components are borrowed, but were formed in English, e.g. ‘Microorganism’. For these I was thinking of creating an additional category for them, but I still need a name for it! I have my review in a couple of days time, and will probably be having a meeting with Lise soon after, so this is something that I will definitely discuss. Thank you for the comment!

  • Gerard O'Grady

    Alex, a quick further thought is this. Are you trying to do two things at once? Perhaps you could think of work in these terms (i) what is nominality and (ii) how is it realised in English?

Leave a Reply

Your email address will not be published. Required fields are marked *