Feedback courtesy of Grant Barret reaches me via secure channels regarding a recent post of mine about observing linguistic data in participants of reality-based TV shows. It appears I was a bit cavalier in my overview of the structure of such shows, and I may have implied (unintentionally) that I believe the course of events in any such show is free of the producers’ influence. In light of Grant’s response, I should (a) acknowledge the influence of producers and (b) comment on the degree of scriptedness in reality TV.

The discussion emerges from the following description I provided:

…there are prototypical features that distinguish reality-based TV from fictional TV. For example, reality TV is unscripted, and the participants are not actors (except in the celebrity-reality subgenre, where the participants are celebrities, but not “on”, or not in character). Much of the footage is also candid and never re-shot.

Grant provides some counterexamples to this: …some of it is, in fact, re-shot, some of it is scripted, and some of it is push-produced. He explains that “push-producing” is a practice of guiding the sequence of events by asking participants how they would react to a particular situation. This makes it difficult to characterize reality TV as totally unscripted, but my position is that there is still a huge element of unscriptedness in reality TV that fictional TV lacks.

The confounding issue here is the component of contrivance that goes into reality TV. Every show has a predetermined structure, which serves as a framework or skeleton for the participants to work with. The framework is essentially the show’s plot, and it is what critics object to when we use the label ‘reality’. For example, why would you vote people off if you were really stuck on a desert island?. The elimination process, Tribal Council, and immunity challenges are all contrived elements of Survivor. The rose ceremony, overnight dates, and meet-the-parents trips are all contrived elements of The Bachelor. Even in reference only to this basic structure, the show has some predetermined topics of conversation for the participants – they’re going to talk about the challenge, or council, etc. This contrivance, in a general sense, is script.

Now, regarding reshooting, I was incorrect. Survivor has been known to reshoot scenes of a particular type, for example, if the producer wants more footage of two people together than what the cameras caught. Moreover, as Grant explains:

…if a “real” person flubs a line or doesn’t say exactly what the producer wants, then the producer will say things like, “Oh, wait, we didn’t get that. Can you just say that again? The part about what a bitch she is?” And the person will do it: usually cleaner, tighter, and on the subject picked out by the producer.

This is an example of reshooting, and possibly of push-producing too. In terms of scripting it comes close to lines for the participant to repeat, but is just short. The topic of the confessional is scripted, but the word choice (other than “bitch”), syntax, and phonology are all still up to the participant and not the producer. My response to Grant, which I’ll repeat here, is that we need to distinguish topic-scripting from syntax-scripting and phonology-scripting.

Fictional shows are scripted at all levels: the specific words and their order are predetermined. My claim is, reality shows do not have this same high degree of scriptedness. This is (possibly) testable, too; the data points in the Apprentice post were the malapropism mediocrisy, the reduplicated exemption-schmejemption, and the singular-agreement they can go frig themself. I’d be wrong about scriptedness here if the producers had specifically asked for the malapropism, for the schm- form (with the [j]), or for themself instead of themselves.

The upshot is that Reality TV shows aren’t as great a source of data as I had let on. They may highlight a few unexpected gems, such as the ones I commented upon, but not enough to build or burst a theory. The reshooting problem obscures the potential for tracking spooneristic speech errors, and the editing makes frequency data impossible to collect. Reality TV is not a substitute for true natural discourse, so in that sense it’s not the same as eavesdropping on the bus.