Friday, March 28, 2014

Friday key note speakers at Antwerp ECSF Conference

I am sitting here in the entrance hall close to the reception desk of the Antwerp Conference.

Here is my report on the three Friday key note speakers:

The first keynote speaker was Ann Packmann. Her speech was about Exploring the relationship between treatment and causal theory in stuttering. She also presented her model of the moment of stuttering, developed by herself and Attanasio. She started out with talking about how one needs to find the cause for a heart issue and back pain before treating it. Unfortunately, that was not a good example, because many people now recognize the biopsychosocial aspects of these issues. Back pain and heart issue can certainly come from or be modulated by psychosocial causes, and treating the physical cause might not be enough. he then talks about  verbal communication & complexity theory. She correctly says that speaking is the most complex thing humans do and which sets them apart from other primates. She goes into complexity theory
and talks about non-linearity and the butterfly effect. She gives an example of the financial crisis being a butterfly effect. Well, having read 10 000 pages for the three CFA exams on finance THAT KEPT ME FROM BLOGGING FOR 3 YEARS, I found this analogy conceptually a bit fuzzy. But needless to say that stuttering is very complex. She then talks about the interaction between body, brain and mind, which again is vague, but not wrong.

She then lists models and theories such as system control modelling (sensory model,neuroscience, variability), cognitive & linguistic (EXPLAN, etc), and multi-factorial (demands & capacity, dynamic multi-factorial). She rightly talks about the difference between triggers and causes, and argues that we should model the moment of stuttering instead of the causes of stuttering. In the same way that we are not interested in hearing that heat and inflammable materials caused bushfires, but what caused a specific bushfire. I partially agree. I would argue that you need to know both: the general cause and the specific cause.

She goes on to introduce her Packman & Attanasio (P&A) 3-factor causal model of moments of stuttering. For them, the moment of stuttering has three factors: neuroprocessing problem (a necessary condition), triggers (features of spoken language), and modulators (inherent to individual). I asked the first question of the conference to her: is a neuroprocessing problem really a necessary condition, as I find that many instances of stuttering could just be learned conditoned responses. She kind of agreed with me, I think. ;-) She went through the three factors in more detail, but I don´t have time to go through them. Nothing earth-shattering. In general, the model seems to me to be somewhat uncontroversial, but it also seems vague and unspecific to an individual. She then asked about explanatory power? testability? parsimony? heuristic value? I don´t really see how it is testable, but it is rather a collection of tested relationships. She then went on to treatments according to the model, and again it is all a bit vague.

At the end, I also asked: How about sub-types in the model? She said that they are not in there. I find that sub-types is a very very important issue, and must be somehow considered.

Overall, a useful and clearly articulated talk, but it was a bit too vague for me and conceptually not as sharp as I wished for.

After the talk, I had a long and intellectually serious chat with her about causal theories of stuttering, which is quite rare at stuttering conferences.

Martine Vanryckeghem gave her speech on Speech-associated attitude and its broader framework (legacy of Gene Brutten). She talked about Gene Brutten, who recently died, I guess. I don´t really know him, so I cannot comment. She talked about his 2-factor model, and his iniative to create a test on patient´s attitude to their speech. The test battery is called CÀT. I am not going into details here, but it was a long list of testing the reliability in different countries.

The third speaker was Kurt Eggers, the conference organizer. He spoke about his work on temperament, a rather controversially discussed topic. He clearly stated that he is not implying that there are strong evidence that temperament is causing stuttering. I have the vague suspicion that Barry Guitar´s "I believe we hit on something deep" at a previous Oxford conference has lost some of its depth. His line of research to look at where there are differences in aspects of temperament are certainly scientifically justified. He had phenomenally visually appealing slides, and also his speaking manner is very politician-like. ;-) His introduction to temperament was clear and instructive. According to Rothbart, temperament is "constitutionally based, individual differences in reactivity and self-regulation. Constitution refers to heredity, maturation, and experience, where we have the big issue that stuttering IS part of experience, and temperament might be influenced by stuttering. Temperament is the difference in reactivity to the same stimulus, with different degree of self-regulation.  This reactivity can have different aspects: somaticity, autonomic, cognitive, and neuro-endocrine.

Kurt talked about different studies that he conducted to study different aspects of temperament. I cannot go into details, but he found differences between the two groups. I am always skeptical by this line of research. For example, some tasks have shown that pws actually perform better in some aspects in visual attention, for example. I would be very surprised. I asked the question and suggested that the experiment somehow failed is a simpler explanation. I further asked about a replication of the results, and he said that no-one had done it, but that other groups had similar results. Replication is so important, but no-one does it. After the talk, I spoke to him more. I also expressed my concern that their control group is actually too clean and elitist, because they took out all kids with issues (e.g. speech impairment). So the control group is NOT a good representation of the population, but free of relevant disorders. Would we not expect them to even outperform a control group that has not been cleaned of disorders? So we should expect a slightly better performance of the clean control group. We also talked about subtypes.