css.php

5 Replies to “The Distribution of the Middle Voice with Psych Verbs in Heritage Spanish (Becky Gonzalez)”

  1. Thanks for the talk, Becky!

    What do you see as some potential methodological implications of your finding that participants (HSs and Spanish-dominant speakers) perform differently with different individual verbs? How should we take that into account during experimental design?

  2. Hi David,

    This is a great question!

    I think this finding is important, although not surprising. In cases like these, where lexical semantics are relevant to structure, I think there are a few things we can do (or keep doing) to take this into account. The first is to include a variety of verbs. Of course there is always a delicate balance of how many verbs to include vs. how many items of each verb, etc., but by testing more verbs we are able to look at both aggregate and individual verb data, which allows us to look more closely at these types of differences in the results.

    Other aspects of experimental design should also depend on the RQ. If we are interested in attainment and want to see what speakers do with verbs they are familiar with, it always helps to use more high frequency verbs. (Of course, as I’ve highlighted here, just because a verb is ‘high frequency’ does not mean every individual speaker will have encountered it frequently in the input, but we can aim for verbs they are more likely to have encountered.) On the other hand, if we are interested in looking at transfer and want to see what speakers will do with a verb they haven’t encountered before, it might make sense to use low frequency verbs. (I could even imagine a more manipulated task where nonce verbs are used for this.)

  3. Very nice analysis of individual trends by verb and subject, and very key in these data. Tough topic too, because of what you mention at the end, with all the polysemy of “se”. I was thinking that the verbs “fascinar” and “entusiasmar” are used a lot in an attempt to translate frequently used verbs in English that are cognates (fascinate) or do not have an easy translation into Spanish (get excited would not translate well with the cognate, excitarse). So I always had the impression that these two verbs in particular were definitely overused by the L2L and some SHL (only a corpus study could shed some light on the truth of this impression). This is all to say that I think that overuse could be related to how they behave differently in how they are perceived. Finally, I think the median rather than the mean would be better for your data, since it is ordinal rather than ratio-scale (and it is a particularly short Likert scale, only 4 points). Very interesting, I look forward to its publication.

    1. Hola Irene,

      Thank you very much for your comments.

      I particularly appreciate your observation about the use of ‘fascinar’ and ‘entusiasmar’ because I had not considered their use (or ‘overuse’) by L2 or HSs. That is a great point and highlights the fact that lexical frequency, as measured by our typical metrics, doesn’t necessarily reflect the frequency of use among the speakers we are testing. It also brings up the idea of lexical equivalents, which are elusive in some cases, like that of ‘entusiasmar.’ During my attempts to provide a a suitable English equivalent for this verb, the feedback I’ve received from different bilinguals has varied. This is definitely a challenge in lexical semantics.

      Thanks again!
      Becky

  4. Great presentation, Becky! I love the middle voice, and I also really like the idea of using spoken stimuli instead of (or in this case, along with written, particularly for heritage speakers.

    Regarding the methods:
    Did you allow participants to repeat the sentences, at all, and if you did, was there a limit to how many times they could listen?
    Also, did you choose to implement any discounted practice items at the beginning to get them oriented?
    Finally, what type of software did you use to implement this study, and was it conducted in person or remotely?

    For the analysis, did you run or consider any statistical test?
    Thanks!

Comments are closed.