Friday, December 28, 2007

Towards a Coherence Theory of Silliness


A number of recent papers have made considerable progress towards giving a full account of silliness. The concept of silliness is a sorely neglected topic in the history of philosophy, and none of the major philosophers have so far written treatises on the matter. Some people look back to Hegel and Jacques Derrida, whose collected works may be regarded as extended meditations on the topic, as pioneers in the field. But most people view it as unfair to regard these writers as “philosophers”; and there is some dispute, even among leading writers on silliness, about whether or not the writings of such people as Hegel and Derrida really do fall under the category of “silliness”; with some commentators regarding the related concepts of “artful nonsense” and “gibberish” as more appropriate in those cases.

The merit of those views, however, is not the subject of this paper, and nor are any of the other new and interesting questions relating to the history of silliness in philosophical writings. Rather, I am concerned here with the concept of “silliness”, aloof from any historical considerations. In particular, I will elaborate upon a particular account of silliness, which I will call the “coherence theory of silliness,” and which I have mentioned briefly in an earlier paper. The core of the coherentist view is that the coherence of a potential silliness-set is a necessary condition for that silliness set to be an actual silliness set. This view may be contrasted with, and has been attacked by proponents of, the “cohesion” theory of silliness. This view holds that coherence is not a necessary condition for silliness, and proposes its own necessary condition in the place of coherence. The key difference between the replacement condition (“cohesion”), and the “coherence” condition, is that the former emphasises the intrinsic relations between silliness scenarios, while the former emphasises the extrinsic relations between silliness scenarios. In this paper I will first respond to some objections to this view that have been put forward by Jones (2006) and Andrews (2006); and will then use my comments upon those objections to motivate some refinements to my own view, which will include a distinction between local and global silliness.

Silliness Sets and Silliness Scenarios

It is almost unanimously agreed that the most useful unit of analysis for the concept of silliness is the “silliness scenario.” The canonical definition of the “silliness scenario” was given by Jones in his pioneering 2005 paper. Later papers have made some refinements upon Jones’ account, but they are of a highly technical nature, and the essence remains the same. For Jones, an event E is a silliness scenario if and only if the following conditions hold:

E must be apprehended by at least two humans agents if the perpetrator of the scenario was a human; and at least one human if the perpetrator of the scenario was not a human. (Discussion about this condition has centred mainly around the problem of zombie perpetrators, solitary silliness, and delayed apprehension. All of these issues warrant further investigation, but they do not pose any real problems for the interaction condition).

(2)Non-Cognition: the agents must apprehend the silliness of E without any cognitive engagement in that event. The agents may, of course, be engaged cognitively with aspects of E that do not actually give rise to any silliness; the agent may, for example, apprehend cognitively some of the intellectually involved parts of a joke, but still apprehend the silly parts without using the cognitive faculty. (This has proven to be the most troublesome condition. There has been considerable discussion about the kind of “cognitive engagement” that is appropriate to this condition, with some advising an abandonment of “cognition” altogether, and settling for a more mild condition; most of these accounts make some use of Davis’ notion of “cognitive relaxation.” Some more recent accounts have investigated the notion of “cognitive tension”, asserting (rightly, I think) that the distinctive features of silliness, which Jone’s (2) gestures towards, is not the absence of cognition, but the tension between what is apprehended using the higher cognitive function, and what is appreciated in the silliness event. If there is some disagreement, however, about just what (2) is gesturing towards, all writers agree that it is gesturing towards something, and that (2) or some variant of (2) is essential to any account of silliness.)

E must be “detached” from any non-silly event. This condition is designed to capture the thought that an event is not usually regarded as silly if it has any genuine real-life consequences (except whatever immediate emotional responses it might elicit from the apprehending agents [see condition (4)]). The silly event must be absurd, free-standing, trivial. As with (2), there is not much doubt that (3) points towards a distinctive feature of silliness; but, as with (2), there is some debate about how to work out the details. In particular, there is debate about what sort of “detachment” is required (logical, topical, psychological, or a combination of those three); and about just how strict the “non-silly” requirement should be. With regards to the second point, some writers prefer to relax the “non-silly” requirement, and replace it with a requirement that is based on “non-funny” events, or even on “non-fun” events.)

(4)Funniness. Clearly the silly event must be amusing: it must elicit a light-hearted response from the agents. The main point of contention surrounding this condition is to do with the relationship between it and the other three conditions; in particular, whether or not this condition is independent of those conditions. Most writers agree that the silly event elicits a different emotional response from the funny event (or else there would no real basis for distinguishing between the two kinds of event). But some hold that this difference should be attributed to the different sort of funniness that inheres in the event; while others hold that that silly events and funny events share the same sort of funniness, and that the difference resides only in the other properties of the silly event (inconsequence, non-cognition etc.) that are absent in funny events. Others (including myself) hold that the debate between these two positions is meaningless.

These, then, are widely accepted as necessary and sufficient conditions for an event to constitute a silly scenario. These conditions may seem irrelevant in this paper, since this paper is concerned with the notion of a silliness set, rather than that of a silliness scenario. But some of those conditions are relevant to the objections that have been levelled against the coherence theory of silliness, and also to my response to those objections. I will move onto those objections, and my response, after outlining the notion of a “silliness set”, and briefly considering the “cohesionist” approach to silliness sets.

Silliness scenarios need not occur in isolation. According to condition (3), silliness scenarios must be detached from any non-silly events; but of course any particular silliness scenario does not need to be detached from other silliness scenarios. In fact, some of the richest and most interesting silly phenomena emerge only out of silly scenarios when they are considered together, and not when each one is considered on its own. Of course, not any collection of silliness scenarios will do. We must consider silliness sets to be collections of scenarios that are related to each other in some substantive way; and the disagreement between coherentists and cohesionists is over just what sort of relation must hold between a collection of silliness scenarios, before that collection can rightly be considered as a silliness set.

Coherence and Cohesion of Silly Sets

On the cohesionist view, the right relation is one that holds between the parts of each scenario in the set, and not the whole scenarios. To work out, from some collection of silly scenarios, which ones constitute a silliness set, each scenario must first be divided into a number of “silly elements”, and then the silly elements from all of the scenarios in the collection of candidate scenarios must then be brought together into a “silly group.” The elements in the silly group are then considered together, irrespective of the silly scenarios from which they were derived. Precisely, they are considered in respect of the richness if the relations that hold between them; this gives rise to a “cluster” of silly elements, all of which enjoy many connections with one another. A scenario belongs to the “silliness set” if and only if all of its silly elements belong to the silly cluster.

The cohesionist view may be contrasted with the coherentist view, according to which the relations to be considered are the relations that hold between the silly scenarios as wholes. There is some temptation to think that the coherentist and the cohesionist views are not in competition at all, but rather that they are equivalent. But this is not the case. That is, there is a genuine question about which of these views gives the right necessary condition for status as a silly set.

To see this, one might consider the analogy of sets of novels (it does not matter if they are silly novels or not). One might take a coherentist view on the criterion for set membership of novels: one consider each novel as a whole (the theme, say), and then determine set membership on the basis of similarities between the themes of the candidate novels. Or, one could adopt the cohesionist view, and hold that the basis for set membership should be the topics of the chapters in the novels: the novels that get into a set should be ones whose chapters are closely related. Now, it is clear in this example that a collection of entities may satisfy the coherentist criterion for set status, without satisfying the cohesionist criterion. This would be the case if the “theme” was an emergent property: one which does not manifest itself in each chapter taken individually, but only when they are considered together. In such a case, it seems likely that the themes of a group of novels may be very similar, but the chapters of each have no special connections between them. A similar situation could plausibly give rise to a set of novels that satisfy the cohesionist criterion, without satisfying the coherentist criterion: the set of novels may have chapters that are very closely related, but which give rise to themes that have no special connection between them. Hence, the choice between the coherence criterion and the cohesion criterion is a genuine choice: one will end up regarding different silly collections as silly sets, if one chooses one of these criteria over the other.

The Cohesionist Attack:

Jones (2005) asks us to consider the following “sillygism”, which he claims to be sound. (In Jone’s terminology, which I will adopt here, a “sillygism” is a list of silly scenarios that have been put forward as a candidate for a silly set. If the collection does indeed turn out to be a silly, the list represents a “sound” sillygism; if not, it is an “unsound” sillygism. If one or more of the scenarios in the silly set are not in fact “silly”, but the scenarios nevertheless satisfy some criterion for set-hood, then the sillygism is “valid” but not “sound.”) Consider the following set of three sillyness scenarios:

1)A man, M2, walks into a bar, B, with a bucket of water, BW in his hand, H. M1 takes B and throws it onto another man, M2, who is sitting at a table talking to a penguin, P1. M2 looks down at his shoes, smiles and utters the statement S1:

S1: “If you do that again I’ll eat your shoelace.”

2)M2 walks into another bar with an Irishman, I, an Englishman, E, and an elephant, EL. EL sees P1 sitting on the bar drinking vodka. E goes up to the bar and orders a Guiness. E goes up to the bar and orders a bar-maid. P1 utters the statement S2

S2: “Throw me a bone, Jim, there’s a shoelace on my soup.”

3)EL walks into a corner dairy, CD, and asks for a penguin. The shop-keeper, S, responds with S3:

S3: “I’m sorry, we’ve run out of penguins. But yesterday the bar-maid delivered a new batch of shoelaces. Will a shoe-lace be sufficient?”

EL replies with S4:

S4: “That should do the trick. Thanks.”

Now, Jone’s claim is that this sillygism, if sound, is a counter-example to the coherentist criterion for silliness sets. For, Jones claims that the coherentist cannot account for the soundness of SA. He claims, that is, that SA fails to satisfy the coherentist criterion. Morevoer, he argues that SA is one of a much larger class of syllogisms that, though sound, do not satisfy the coherentist criterion. If Jones is right, the coherentist account is in serious trouble.

It is hard to deny that SA is a sound syllogism. Some (Pritchard, 2006) have raised some doubts about its soundness, pointing for example to the introduction of the vodka and the Guinness in 2), and the shop-keeper in 3). Some have gone so far as to question the validity of the set, pointing in particular at the “funniness” condition for silliness. In Pritchard (2006) we find the claim (for example) that “while SA3 might raise a giggle for some people, SA2 is pretty lame. And SA1 is scandalous. Quite simply, SA is not funny.” But Pritchard’s intuitions about silliness have been shown to be deviant in other cases; and most philosophers, including a number of logicians, agree that SA is funny.

The novel part of Jone’s argument (and the part I wish to attack here) is his claim that the coherentist criterion cannot account for the soundness of SA. Before giving my response to Jone’s claims here, I will rehearse his argument briefly.

To be continued, given further inspiration. In the meantime, consider this silliness scenario (especially the sub-scenario in the 30-60second range).