Eye Contact, Ear Contact, and Autism

Of all the topics I’ve ever dealt with on this blog, the one that has continued to attract the most attention is children’s eye contact. Hundreds of people daily read blog posts on this topic, and when I look at the questions that are asked and bring readers to this site, the great majority of them are about eye contact. People know that unusual use of the gaze is characteristic of autism and become concerned when a child does not look at the adult’s eyes in the way they might expect.  

Unfortunately, that justified concern is all too often coupled with unrealistic expectations about the development of eye movements and eye signals as part of communication. Some young parents become very anxious when a baby of a few weeks old does not readily gaze at someone’s face, or even when a nursing baby closes the eyes while sucking. It seems that people expect very young babies to use the eyes in almost the same way that adults do, rather than understanding that communication with the eyes--  and even vision itself—are not completely developed at birth. It seems to be forgotten, too, that human beings have communicative modes other than eye contact, and that failure to develop speech is an outstanding characteristic of autism.

It is not possible to diagnose autism from a baby’s use of gaze at a few months of age. but could it be possible to see the beginnings of other features of autism at that early stage? Babies don’t talk, of course, but might there be something about their sounds and their responses to others’ sounds that could predict development along atypical lines? A group of French investigators, Julie Brisson and her colleagues, looked at 13 babies who were later diagnosed as autistic (LDA) and 13 others who were typical in their development (TD). All the babies’ families had taken home movies that recorded the sounds made by the babies and their mothers in the first six months of life. (Of course they did not know at the time which babies would develop in which way; the investigators asked for the movies taken earlier, after the children had been diagnosed with or without autism.)

An interesting part of this study was the examination of the mothers’ speech to the babies, as well as of the babies’ own sounds. No matter what language a person speaks, he or she talks to babies differently than to adults. The type of speech used with babies is sometimes called “motherese”, but is also referred to as infant-directed talk (IDT) to differentiate it from adult-directed talk (ADT). IDT uses a higher pitch and more intonation or change in pitch, as well as repeating phrases or words more frequently than in ADT. Babies are typically very interested in IDT and will pay attention to someone speaking in that way, even though they will ignore the same person speaking ADT. It seems to be the case that adults are “taught” to use IDT to some extent by the responses they get from interested babies. If a baby was not interested in IDT (as might be the case for a baby developing along an autistic pathway), he or she might not reward an adult by attentiveness, so the adult might quite unconsciously not use IDT so much.

As babies develop through the first months of life, they are already beginning to change the sounds they make toward a more speech-like pattern. Because autistic children often speak late or have trouble developing speech in other ways, it’s possible that the LDA babies already showed unusual sound production patterns before they were six months old.

Brisson and her colleagues found that even in the first six months the sounds made by LDA babies were more monotonous and less complex in their sound patterns than those made by TD babies, although they were similar in pitch and duration. The mothers’ speech to the LDA babies used shorter utterances than speech to the TD babies, which might have been because when the babies failed to respond to talking, the mothers tried following up with a shorter message (like the baby’s name). But with little feedback or reinforcement of their efforts, the mothers may have become less likely to talk to the LDA babies.

Could we say, then, that the mothers somehow caused the infants’ autism by failing to talk to them enough? This is not likely to be an accurate conclusion, because there seems to have been a pre-existing characteristic of the baby that discouraged the mother from talking, or at least failed to encourage her as more typically happens. In either typical or autistic development, the development of communication is transactional--  each of the pair affects the other, and the way they do so changes over time. However (and this is my speculation, not Brisson’s), it’s an interesting question whether the LDA babies might have developed more typically if the mothers’ speech to them had been “artificially” increased by training, rather than by the usual rewarding baby responses. In a 2012 article in Science, Beaudet speculated about the possibility that the genetic factors in autism may make LDA babies need different nutrients than those appropriate for TD babies, and if they received an appropriate diet the LDA babies might develop more typically. Beaudet’s idea has not been supported by much evidence so far, but it is an intriguing one—and it does suggest the idea that there could be a parallel in terms of a “speech diet” designed for LDA babies.

Let me just emphasize that this discussion is by no means intended to blame mothers of autistic children. Transactional processes being what they are, a mother’s speech behavior is as much shaped by her baby’s responses as the baby’s sound production is shaped by the mother. It’s possible that LDA babies don’t have the ability to “teach” their mothers how to talk to them as TD babies do. Talking IDT is not something that’s easy to do voluntarily--  just for fun, try to talk to another adult or an older child as if you were talking to a baby. But if it turned out to be a good idea to help mothers of LDA babies do more of this (which we can’t really tell from this small study), we could certainly find ways to encourage them.

ADDENDUM: I forgot to give the citation for the Brisson article. It is Brisson, J. et al (2014). Acoustic analysis of oral productions of infants later diagnosed with autism and their mothers. Infant Mental Health Journal, 35, 285-295.