Natural language research announcements

[Margaret Fleck <mfleck AT cs.uiuc.edu>]

I don't usually forward lpbb talk announcements to nl-uiuc.   However, there's
a glitch (see below) with this week's lpbb mailing which may cause some of you
who normally get lpbb mailings not to have gotten this one.

Cheers,

Margaret

Scott Jackson wrote:
> Dear LPBBers,
>
> First, apologies if you're getting this message unsolicited.  My usual 
> routine is to run these
> messages through Judy Allen, but she's out of town this week, and I'm working 
> off an older list
> of addresses.  So PLEASE spread the word about the talk this week, in case 
> someone who would like
> to come hasn't gotten the message...
>
> In any case, I'm happy to announce that we have an outside speaker this week 
> in the LPBB: Peter
> Richtsmeier, a postdoc at the U of Kansas.  His talk is titled 
> "Four-year-olds Generalize
> Phonotactic Probabilities from Ambient Speech", and the full abstract is 
> below.  The talk is in
> the normal time and place, 12:30 this Thursday in Beckman 4269.
>
> Additionally, Peter will have time to meet with interested people on Thursday 
> and Friday.  If
> you'd like to meet with him, just drop me an email about when and for how 
> long, and we'll work it
> out.
>
> Hope to see you Thursday! -scott
>
> TITLE: "Four-year-olds Generalize Phonotactic Probabilities from Ambient 
> Speech"
>
> ABSTRACT: Through a series of four experiments, I will present data that 
> inform our empirical
> understanding of what phonotactics are and how they are acquired. The 
> methodology is similar to
> studies of infant perception: children are familiarized with a set of words 
> that contain either a
> few or many examples of a phonotactic sequence (Gerken et al., 2006). Unlike 
> infant studies, the
> participants are four-year-olds, and the test involves producing the target 
> phonotactic sequences
> in new words. Because the test words have not been encountered before, 
> children must generalize
> what they learned in the perceptual familiarization and apply it to their own 
> speech. By
> manipulating phonetic (talker variability) and phonological (word-type 
> frequency) characteristics
> of the familiarization items, we can gain insights about the nature of 
> phonotactic learning. With
> respect to the results, I will first show that the well-studied correlation 
> between phonotactic
> probability and production acc uracy in child speech can be at least partly 
> attributed to
> perceptual learning, rather than a practice effect attributable to repeated 
> articulation. Second,
> I will show that perceptual learning is a process of abstraction and learning 
> about those
> abstractions. It is not about making connections between stored, unelaborated 
> exemplars or about
> organizing pre-existing symbolic knowledge. I argue that a model 
> (Pierrehumbert, 2003) which
> learns abstract word forms from direct phonetic experience, then learns 
> phonotatics from the
> abstract word forms, is the most parsimonious explanation of phonotactic 
> learning. Furthermore,
> this model has already been extended to explain many facts about phonological 
> acquisition.
>
> REFERENCES: Gerken, L. A., Goffman, L., Carter, A., Bollt, A., Bruner, A., 
> Fava, E., et al.
> (2006). Statistical frequency in perception affects affects children's 
> lexical production.
> Unpublished manuscript.
>
> Pierrehumbert, J. B. (2003). Probabilistic phonology: Discrimination and 
> robustness. In R. Bod,
> J. Hay, & S. Jannedy (Eds.), Probabilistic Linguistics (p. 177-228). 
> Cambridge, MA: MIT Press.