Sometimes the main thing I hope to accomplish with a blog post is to turn you on to something that’s fun to read. This is one of those times. If you’re at all a verbal behavior person, check out this easy-to-digest article with a linguistics theme from Atlas Obscura. Topic: the “Crispy R,” a minority pronunciation of the letter R that is hard to describe, but you’ll probably recognize it in a linked video that documents its pronunciation (or, go directly to this YouTube video, but be forewarned: the speaker demonstrating the sound is super annoying).
If the Crispy R is an onion, it has all kinds of layers. Here are a few.
Linguists have done a precise task analysis of what physical actions are required to produce the Crispy R. In general, it’s not easy to form an R, which is why, developmentally, this sound tends to be one of the last to be mastered by children (think baby talk: e.g., “widdle wabbit”). Speech therapy focuses so often on teaching widdle wascals to wender the sound wespecably well that there are book-length treatments of this single problem.
Part of the difficulty with the R sound is that there are multiple ways in the mouth to create it. Some of the physical variants are virtually indistinguishable to the human ear. Where the variants can be distinguished, however, a robust generalization gradient seems to apply, in that listeners recognize several nonidentical R sounds as, well, R sounds. The Crispy R, for instance, has a unique “mouth topography,” and it sounds different from other Rs, but when you hear it, you still know it’s an R.
This doesn’t mean, however, that all distinguishable Rs are created equal. Some may convey higher social status than others. An historical example is the “non-rhotic” R, which is essentially not voiced. Here’s a quick video pronunciation guide (Be aware: This speaker is incredibly annoying too. Save yourself some irritation, and view only between 0:53 and 1:44 — though trust me, it will feel a lot longer than 51 seconds. Gosh, are all linguistics nerds this insufferable? Hmm. If so, that could prove useful… By comparison, behavior analysts should seem warm and fuzzy… So maybe whenever we meet with interdisciplinary teams or members of the public, we should bring a linguist along, to harness the contrast effect…. Ah, but I digress.).
You can hear (or, more precisely, not hear) the non-rhotic R in many pre-World War II American movies, in which American actors employ vaguely British-sound cadence and pronunciation. The British connection is key, as non-rhoticity is endemic to British English. As the article notes:
In American English, that non-rhoticity was, from the colonial era up to the early 20th century, considered prestigious: It was associated with the wealthy port cities of the Northeast that had extensive contact with Europe. (Think of FDR’s “the only thing we have to feah is feah itself,” or basically anything JFK ever said.)… Non-rhoticity vanished from the upper classes as the United States overtook England as a world power.
Thus, when Britain was held in high esteem in America, a lot of people used the non-rhotic R. When Britain waned in prestige, so too did British-sounding Rs.
As for the current status of the Crispy R, perhaps all that need be mentioned is that a Kardashian uses it and, although hard data still are scarce, it may be gaining in popularity. This could represent the early stage of a phonological shift (society-wide change in pronunciation), which occurs all the time in language. Phonological shifts are part of why, for instance, Shakespearian English from the 1500s can be hard for a person on today’s streets to understand, and Middle English (~1100 to 1500) sounds like a whole other language (it’s not just that words have changed; the way we pronounce them has as well). In American English, the past 150 years have seen a gradual shift in vowel pronunciation such that “black” sounds increasingly like “block.”
With Kardashians in mind, it’s interesting to note that phonological shifts tend to be driven by women, often young women (e.g., see here and here and here), which is a whole fascinating topic in how societal forces intersect with verbal behavior — one for which behavior analysis currently offers no ready explanation. But for present purposes, suffice it to say that if a Kardashian does it, a lot of other people probably will too.
Skinner’s Influence
When Skinner wrote Verbal Behavior, language scholarship was firmly in the grip of linguistic structuralists whose interest, for various reasons, was in surface features like categories of words (nouns, verbs, adjectives, etc.) and grammar and syntax. Skinner’s great contribution was to highlight the function of language, pointing out that its topography could be misleading (consider the “mand in tact’s clothing” or the “magical mand“). In contrast, Skinner’s primary verbal operants are, speaking loosely, categories of utterances with shared function, not specific words or phrases. Grammar and syntax aren’t taken up formally in Verbal Behavior until Chapter 13, where they are treated as emergent effects of functional processes. The other great contemporary influence on behavioral accounts of language, Relational Frame Theory, continued the functionalist tradition. Within the context of relational frames, words and other verbal responses can control behavior in ways that deviate from dictionary denotations.
Skinner wasn’t naive. He understood that verbal topographies are critical to speaker and listener perceptions of what each other are “saying.” As he wrote in Chapter 2 of Verbal Behavior: “What is needed [for a proper analysis of verbal behavior] … is a unit of behavior composed of a response of identifiable form functionally related to one or more independent variables. In traditional terms we might way that we need a unit of behavior defined in terms of both ‘form and meaning.'” Yet, in the eyes of behavior analysts, Skinner may have succeeded too well in vanquishing structuralist traditions in linguistics. For instance, check out Postscript 2, which lists article titles from the last 5 issues of The Analysis of Verbal Behavior. You’ll find little obvious attention to verbal topography.
What Gets Left Out
Thus, although our functional approach has yielded lots of benefits, we may have misplaced a baby while changing out the bathwater. As concrete evidence of this, consider: Applied behavior analysts do a ton of work to establish verbal repertoires, but if they don’t know how to teach a proper R sound, how far can this process really go? To serve its function, verbal behavior has to sound like verbal behavior (i.e., speaker and listener need overlapping topographical repertoires), so the knowledge and skills of speech and language therapists seem integral to what many applied behavior analysts work on every day. There is no attention to shaping verbal topographies per se in the standards for training applied behavior analysts, and for whatever reason, there appears to be limited cross referring between the two disciplines, so (with apologies to the folks mentioned in Postscript 1) one can only conclude that applied behavior analysts are mostly winging it when it comes to shaping up the form of verbal behavior — which certainly raises questions about this item from the Behavior Analyst Certification Board’s Ethics Codes for Behavior Analysts:
1.05 Practicing within Scope of Competence
Behavior analysts practice only within their identified scope of competence. They engage in professional activities in new areas (e.g., populations, procedures) only after accessing and documenting appropriate study, training, supervised experience, consultation, and/or co-treatment from professionals competent in the new area. Otherwise, they refer or transition services to an appropriate professional.
But rather than casting aspersions at people who are doing the absolute best they can with the training they’ve been provided, let’s view verbal topographies as an opportunity to expand the focus of verbal behavior research and practice.
Skinner was right that, within a verbal community, topographical norms of language are not very interesting. They are the end product of social-verbal forces that cannot be reconstructed or understood simply by describing topography. But two things are very interesting,
First, topographical features shared within one verbal community may evoke selective responding from member of other communities. To get a feel for this, try carrying an Appalachian (“hillbilly”) accent into a high-powered job interview. I suspect Relational Frame Theory could have a lot to say about the biases, positive and negative, that are associated with various verbal topographies.
Second, changes in topographical norms (phonological shifts) reflect dynamic forces that seem very much compatible with the science of behavior as we know it. But as a scientific focus this is uncharted territory for behavior analysis. I’m guessing that this is the case partly because the relevant forces operate at many scales. Individual speakers are of course affected, but for the norms of a verbal community to shift there may be more at work; think of this as a sort of social psychology of verbal behavior.
Once we acknowledge the importance of phonological properties of verbal behavior, there’s even more to consider. For instance, in tonal languages phonetics as we think of them in the Western world combine with changes in vocal pitch to convey “meaning.” This is complex stuff for which no behavioral analysis currently exists (as far as I know, anyway). Some aspects of vocal tonality may operate more or less as do standard phonetics — that is, “meaning” emerges from shared verbal repertoires with respect to arbitrary topographical aspects of language. It’s not clear, however, that all topographical features are strictly arbitrary. It’s interesting that, in child development, across all cultures, vowel sounds emerge before consonants, which emerge in a more or less standard order. This suggests a strong influence of physical development over the earliest development of verbal capabilities.
Later in life, strategic use of intonation and vocal rhythm can produce profound listener effects that at least supplement other verbal effects, and may even be independent of them. One of my most vivid undergraduate memories is of listening to an English professor read Beowulf in Old English. I didn’t understand a word of it, but the mellifluous rise and fall of his rich baritone kept me captivated. At times I felt physical tingles; at other times I choked back tears. [And I’m not one to readily emote; for instance, I didn’t cry at my father’s funeral; when I fractured my hip in a bike crash; or when WVU relented and gave me a PhD.]
I cannot describe exactly what my instructor was doing to achieve these effects, but religious leaders, politicians, salespersons — people whose success depends on influencing others — have a similar skill set. In short, there are phonemes, and then there are the vocal tricks we use to make them sound like poetry. By the way, the best known and most widely applied manifestation of this skill set is in singing; Ruth Anne Rehfeldt and colleagues (2021) have offered a start on making sense of how music affects us. See also a cool study by Martins and colleagues (2023) on how musical notes participate in stimulus classes. But there is much more to be learned.
Speakers and Listeners and the Future of Verbal Behavior Studies
Thinking about rhythm and intonation drives home how we arrived at where we are, with little to show in our analysis of verbal behavior regarding some fascinating aspects of communication. In defining his agenda in Verbal Behavior (Chapter 2), Skinner spoke directly to formal, non-phonemic properties of speech such as “energy-level,” “speed,” and “repetition.” But to Skinner these were clues to what controlled speaker behavior. The phenomena I’ve mentioned here are, to a large degree, defined in terms of effects on listener behavior — how attentive they are to speaker behavior, how they feel about the speaker, other emotional responses, how likely they are to change their behavior in response to speaker behavior, etc. This is the great blind spot in Verbal Behavior, with its limited attention to language as a tool of social control (to me, this is odd since listener control is directly relevant to an understanding of speaker behavior, as listener behavior is a powerful form of momentary feedback for the speaker). Relational Frame Theory does better in emphasizing the role of the listener, although to my knowledge it has not yet been applied to problems like those mentioned here.
Within behavior analysis, the history of verbal behavior studies has seen two major chapters. First Skinner laid out a functional agenda and defined some elementary verbal operants (speaker responding). Then Relational Frame Theory gave us tools for better understanding the “symbolic” nature of verbal behavior (listener responding). Maybe, in order to expand our science and more fully account for the range of verbal phenomena in the everyday world, it’s time to open a third chapter in which the focus is on topographical features of language, the behavioral forces that influence them, and the effects they have on listeners. That might just give us the complete science of verbal behavior (“form and meaning”) for which Skinner said we ought to strive.
Postscript 1: Credit Where Due
I don’t mean to give the impression that behavior analysis has entirely ignored issues like those discussed here. A relative handful of people have explored the interface between the applied behavior analysis and speech-language professions. For instance, check out ABAI’s Speech Pathology ABA Special Interest Group.
Postscript 2: Hegemony of Function
Article titles (most recent 5 issues) from The Analysis of Verbal Behavior (excludes a 2021 series of Jack Michael remembrances). Note the limited attention to verbal behavior topography.