[From Bruce Abbott (2015.09.27.11:15 EDT)]
Delayed auditory feedback has a profound effect on speech production within a certain range of delay. (In the cited article a delay of 200 ms produced the greatest disruption.) I’ve experienced this myself. One has a strong tendency to start syllables over again and to stretch them out, effects that showed up in the cited study as an increase in the number of pronounced syllables and an increase in their duration. It’s a dramatic demonstration that speech output is under negative feedback control.
Speech production involves a much more complex mechanism than the simple control systems we usually discuss on this list.  We have to generate the proper words in the right sequences, with the correct grammatical endings, accents on the proper syllables, and so on. When we speak, we monitor our speech output. We actively control its perceived loudness and certain other continuous properties such as its tempo. But speech production also involves producing the proper sounds one after another with the correct timing. Under delayed auditory feedback, it feels as though we are now attempting to produce the next sounds in the sequence before the previous ones have been completed. The mechanism controlling the perceived sequencing of these sounds detects an error and tries to correct it by starting the current output over beginning with the sound just heard, or by slowing the rate of production so that the current output is not perceived to be overlapping the previous output just now being heard.
In the cited paper, the authors state that the sound of one’s own voice provides feedback to the system that produces vocal output, but that raising one’s voice when it is being masked by strong “pink� noise provides an example of feedforward. In the latter case, it is assumed that the participant raises the voice in order for an audience to hear it better.  Supposedly, an autistic person would be less able to understand the listener’s need for the speaker to talk more loudly in a noisy environment, and therefore would be less likely to raise his voice in the strong pink noise condition.
The authors call this adjustment of voice intensity “feedforward� because the listener’s ability to hear the speaker is not fed back to the speaker; rather, the adjustment is done without feedback from the listener but instead based on prior experience. “Feedforward� in this context means “open loop� with respect to the effect of the adjustment on the listener’s comprehension of the speaker’s speech.
A full analysis of all this from a PCT perspective would be a rather daunting task requiring more knowledge of the mechanisms involved than I possess. However, this so-called “feedforward� process can be handled quite easily. Under ordinary conditions we soon learn to vary our speech intensity according to the feedback we receive from listeners (e.g., “Speak up, I can’t hear you!). We learn the conditions under which a listener is likely to have problems hearing us (e.g., strong ambient background noise) and reorganize so as to vary our speech intensity according to those ambient conditions. When conditions arise under which, in the past, listeners have had trouble hearing us, we raise the reference level for the loudness of our vocal output. There being no complaints from the listener, we perceive no error in the upper-level control system that raised the lower-level’s reference for voice intensity so as to make communication possible given the ambient noise. That’s ordinary negative feedback at work.
Bruce
···
From: Warren Mansell [mailto:wmansell@gmail.com]
Sent: Sunday, September 27, 2015 4:01 AM
To: csgnet@lists.illinois.edu
Subject: Delayed Auditory feedback effects versus external noise?
Anyone got a PCT rather than Bayesian explanation for this?
http://journal.frontiersin.org/article/10.3389/fnhum.2015.00510/abstract
Individuals with autism spectrum disorder (ASD) show impaired social interaction and communication, which may be related to their difficulties in speech production. To investigate the mechanisms of atypical speech production in this population, we examined feedback control by delaying the auditory feedback of their own speech, which degraded speech fluency. We also examined feedforward control by adding loud pink noise to the auditory feedback, which led to increased vocal effort in producing speech. The results of Japanese speakers show that, compared with neurotypical (NT) individuals, high-functioning adults with ASD (including Asperger’s disorder, autistic disorder, and pervasive developmental disorder not otherwise specified) were more affected by delayed auditory feedback but less affected by external noise. These findings indicate that, in contrast to NT individuals, those with ASD relied more on feedback control than on feedforward control in speech production, which is consistent with the hypothesis that this population exhibits attenuated Bayesian priors.