Comments

  • Alison Wray

    Manal, what an interesting pair of experiments. I like how you are looking for an explanation for the contrary results by considering the complexity of the underlying processing. The suggestion you make in the Discussion seems worthwhile exploring, and I’d like to add an extra idea.
    So far you have suggested that the advantage that cognates have is their phonological similarity. But suppose that cognates are paired at a deeper level – that is, the nodes are linked because the speaker knows they are cognates. We already theorise nodes being linked semantically – that’s what priming captures. And we tend to theorise phonological priming as more shallow: I hear a sound or see a letter and I start considering everything that matches, gradually filtering out options as I get more information in (this is the cohort model, https://en.wikipedia.org/wiki/Cohort_model).
    As I understand it, you are thinking of your cognates as undergoing this cohort process, but of course they are also semantically linked. Being both semantically linked and phonologically similar already suggests an accessing advantage, but the fact of being recognised as a cognate might be an additional way in which the nodes are linked, or even (if we think in terms of compound and coordinate bilingualism models) compounded, ie one node for both.
    If so, then there are some interesting opportunities in an experimental design, where you separate out aspects of that association, e.g. (if such instances can be found)
    – pairs of words that are both semantically linked and phonologically similar but are not cognates
    – pairs of words that are cognates but are not strongly semantically linked, due to semantic drift in one language or other
    – pairs of words that are cognates but the phonological link is more obscure and so people tend not to realise they are cognates
    Hope this is useful.

  • Manal

    Hi Alison,
    Thank you for your valuable feedback. Your idea is really interesting and worth consideration. The problem with cognates is that there are only a few of them. In each experimental condition, we need around 15 pairs of words and there are less than 7 cognates that might fit into your last suggested condition. I will see if I can come up with a workaround.
    Thank you again for your insights.
    Kind regards,
    Manal

  • Kate Kavanagh

    Thanks very much for your poster Manal, it is very clearly laid out but I am quite lost with all the technical terms! It would be good to have some definition of key terms for non-specialist readers (like ‘cognates’ and ‘selection process’ for example). Cheers!

  • Gerard O'Grady

    Hi Manal, thanks for your description of your experiments. It is fascinating that you have apparently found contradictory results and I like the insight that selection is only competitive if there is a reason (or could this be need?) for such competition. Would this imply that views which either propose that each language is represented in its own network or on the contrary that there is only a single network are too crude? And what is going on is much more dependent on individual experience and need? So what is activated and supressed depends on the results of input and hence will vary between people. As a framework I’m thinking of Edelmann’s neural Darwinian hypothesis. Of course if I’m right the results will be very messy.

  • Manal Alharbi

    Hi Kate,
    Thank you for your comments and your question. The selection process is what we call the situation where bilinguals must select the right word in the right language among several activated words in the lexicon. For example, when English-Spanish bilingual, were presented with a picture of a dog and asked to name it in L2, the two corresponding lexical nodes were activated in the lexicon (dog and perro) among other related nodes. The task as in my experiments requested them to name in L2. So, in this case they have to select the word perro (the right word in the right language). The question under the investigation is that are these two compete for selection or not.
    Cognates are words that are semantically and phonologically similar. They may happen in a language or in a group of languages, for example, telephone in English and teléfono in Spanish.
    Hope this clarify those technical jargon :)
    Many thanks.

  • Manal Alharbi

    Hi Gerard,
    Thank you for your comments and suggestion.Yes, exactly it looks like input type is behind the whole observed competition and suppression patterns.So far I have tested whether proficiency level would influence the selection process and found that it has no significant impact.
    Thanks again for your suggestion and I’ll have a look the Edelmann’s neural Darwinian hypothesis.
    Regards, Manal.

  • Kate Kavanagh

    Hi Manal, that does help clarify, thank you! :)

  • Katharine Young

    Hi Manal! Great poster, thanks for sharing :) I’m interested to know how you measured bilingual fluency in your participants – was it self-reported?

  • Manal

    Hi Katharine,
    Thanks for your comment and question.
    I used their IELTS scores and a self-report assessment.
    Regards, Manal.

  • Michelle

    Nice poster Manal. Thanks to Alison & Gerard for their very useful suggestions, let’s explore these, Manal at our next meeting

  • Manal Alharbi

    Thank you Michelle :)
    Sure, I have already started reading.

Leave a Reply

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