GUEST BLOG by: Einar T. Ingvarsson, PhD, BCBA-D, LBA
Virginia Institute of Autism
The movie Arrival (2016) tells the story of an alien life form visiting Earth for the first time. The Heptapods (seven limbed aliens) land at 12 sites worldwide and are quickly engulfed by hordes of scientists and military personnel. Linguist Dr. Louise Banks (Amy Adams) is recruited by Colonel Weber (Forest Whitaker) to establish communication with the aliens. The Heptapod language turns out to be, well, alien; not like any human language previously known. Colonel Weber’s initial expectations of Louise’s contributions are typical of common misconceptions of the nature of language. He brings her an audio-recording of the Heptapods’ rumblings and expects her to translate their message purely based on auditory patterns. Louise promptly tell him that this is not really how translating an unknown language works. She quickly realizes that understanding of the alien language can only be based on common experiences that they share. Louise and her physicist colleague, Dr. Ian Donnelly (Jeremy Renner), begin establishing communication with two Heptapod representatives. The process begins with common vs. individual identities (i.e., “Louise” and “Ian” vs. “human”), then proceeds to simple actions (e.g., “Ian walks”) and so on. Months of painstaking work eventually lead to a rudimentary common vocabulary, which then leads to … well, I won’t spoil the movie further for those that haven’t seen it. You can watch it on Amazon Prime. It’s well worth it.
Although Louise is an academic linguist and does not use any behavior analytic terms to describe what she does, she sometimes behaves like a radical behaviorist. First and foremost, her approach to translation is in large part based on behavioral function. Her work illustrates one of the major points of B.F. Skinner’s (1957) approach to language: topography or form is meaningless alone; it only derives meaning from association with environmental (i.e., non-verbal) events. By identifying how environmental events control verbal behavior, we can determine the meaning of words and other symbols.
Louise and Ian conduct something akin to tact training with the Heptapods, by pairing actions and items with words and symbols. To account for the vastly different auditory topographies, they augment their communication through visual symbols; written text on a whiteboard for human language (English) and the Heptapod’s corresponding circular visual symbols. Although we only see brief glimpses of how they build the common understanding of a vocabulary in the film, it appears that the process of recombinative generalization may be involved. After establishing tacts of individuals (e.g., “Ian”) and actions (e.g., “walks”), the individual and action are presented together (“Ian walks”). By systematically presenting multiple individuals and actions together, responding under the control of novel combinations is likely to occur without direct training.
Behavior analysts have explored systematic ways to accomplish recombinative generalization through an approach called Matrix Training. For example, one could teach a series of tacts under the control of at least two different elements, such as “red circle”, “yellow square”, and “blue triangle.” This might lead to the individual tacting a red square, yellow triangle, blue circle, and so on, without direct training. Recombinative generalization relies on the existence of individual tacts under the control of environmental objects and events (e.g., “Ian”, “walks”). Behavior analysts have referred to this as minimal response repertoires or atomic repertoires. Each stimulus (in this case a word) controls a distinct behavioral unit. When the stimuli are presented in novel combinations, the corresponding combination of behavioral units occurs. This way, a virtually infinite number of novel combinations of behavioral units can occur without having been directly reinforced or prompted in the past. This contributes to a flexible and generative verbal repertoire.
As a common vocabulary of tacts is established, Louise starts identifying meaningful patterns in the Heptapods’ symbols and vocalizations and meaningful conversations are now possible. These involve asking questions (i.e., mands for information) and answering questions (intraverbal responses). Louise can start conversing with the Heptapods to find out what their purpose on Earth is. The parallels with behavior analytic language interventions might be apparent to some readers. Learning to answer or ask questions without the corresponding nonverbal stimulus control would be like learning a foreign language only through verbal imitation. You could say the words, but you could not make the use of the foreign vocabulary in any meaningful way. However, if you learned to tact objects and actions in the foreign language, a functional vocabulary would be established.
Incidentally, this would not be a problem if the foreign language teacher already knew your language. In that case, the teacher could simply teach you foreign language translations intraverbally (e.g., “in Icelandic, borð means table”). However, Louise and the Heptapods do not share a language. The same is true in early intervention for children with autism spectrum disorder (ASD) who have severe language deficits; by definition, there is no shared language already in place between the interventionist and the child. A child can be taught to answer a question such as, “How old are you?” or “Which Avenger has a magic hammer?”, but without the corresponding non-verbal stimulus control (e.g., being able count up to ten items and tact or identify Thor and Mjölnir), the responses would have little functional importance for the individual. Hence, establishing nonverbal stimulus control from early on is critical for meaningful communication.
In conclusion, whether the issue at hand involves translating alien language, learning a foreign language, or language interventions for children with ASD, meaning is an important outcome. It is important to note that terms such as meaning and understanding are not explanatory terms in behavior analysis, but they can be dependent variables. As such, they can be explained by reference to basic behavioral principles (e.g., stimulus control and Skinner’s analysis of verbal behavior). Behavior analysts should not hesitate to teach the general public how our science can contribute to the understanding of meaning.