Re: A preliminary look at Spoonerisms

Glen M. Sizemore wrote:
Skinner's view of verbal behavior involves a hierarchy of response units. For example words are not necessarily units in any particular utterance, but they often retain some functional significance. The same is true with smaller units. When we are repeating a very unfamiliar utterance that we have just heard, our utterance strongly involves our "minimal repertoire" - units that are part of all of our utterances. The same is true when we attempt to echo a foreign word or phrase (which may not be completely possible because of the difference in minimal sounds that characterize languages).



It is often mistakenly assumed that behaviorism treats all verbal utterances as a chain of responses where words early in the sentence exert stimulus control over later words. In some cases this is largely true as in certain kinds of intraverbals (1-2-3-4 etc.), but this does not characterize other sorts of utterances. Instead, his view was that, for example, some circumstance strengthens some core responses (let's look at tacts for the moment) and then the core responses become temporally organized as a function of contingencies that produce other units that we call grammar and syntax (these are autoclitics). The utterance is often emitted as a complete occurrence, not assembled after "the early parts" are emitted. Skinner felt that Spoonerisms provided evidence that this, in fact, was the case since later parts of the response appeared to exert control over earlier parts. For example we might say "ghouls fold" when identifying iron pyrite in part because the "guh" sound is a fragment of "ghouls" and, of course, "gold." Given that the "ools" in fool's is also made more probable, and "ghouls" is itself likely a unit possessing some probability of occurrence, the Spoonerism occurs. Thus, utterances are built up sequentially, and inchoate portions of the early response exert control over the final product, the sequence of events is not from early words to later words, and doesn't necessarily involve whole words as the only units. Obviously, it involves the autoclitic units that comprise the organization of the basic tacts or mands.



It should be possible to show (at least statistically) that the "erroneous" statements can be controlled by variables that affect the probability of the erroneous substitutions. For example, the response "ghouls fold" may be more probable following exposure to "Halloween themes" or among poker players or those that practice origami. I recall an article I read some time ago (I think it was about verbal slips) in which one of the examples given involved people that were threatened with electric shock (though none was ever really delivered) for errors in identifying briefly-presented written phrases. In this group, there was some elevated probability that "worst cottage" would be reported as the Spoonerism "cursed wattage."




You might as well ask what makes the unexpected, or the deviation from
the expected funny. What indeed is funny, what is the behavior in this
aspect of language and biology.

To answer my own question, i'd say deviation from the expected comes
with a degree of apprehension. The actions of laughter has properties
in it which might be liken to other aspect of flight or fight, the mechanics of laughter involves us breathing deeper, expelling air,
the acidity which comes with work or apprehension. Laughter, before
its social context, could be seen as an adaption of this older
mechanism, as such the joke could be seen as a benign form of stress
and then release. The structure of many jokes takes this form of the
predictable, or the set up and then the jump away from that prediction.
Or else the laughter is based on recognition. sharing an idea, or an
image we all agree with. This is it seems is a variation on the idea
of prediction.

In a round about fashion bring me to your question on spoonerisms,
which it seems to me is just one variation of the standard setup,
the contradiction, the recognition. What in this idea of a joke
would a computer find Puny?
.

Quantcast