Wednesday, 18 March 2009

Why Does Fad = Bad? Velocity, Autonomy and Constancy

It’s a fact that trends end, but most accounts of the uptake of cultural artifacts don’t really get into the end too much. They focus on the dynamics in regards to acceptance or consumption, not really touching on how these very same dynamics can lead to the cultural item being very quickly dropped. Trends, after all, are only as stable as people’s beliefs about other people’s preferences and expectations.

Which brings us to fads. Fads can be thought of as cultural trends – tastes, styles, fashions, attitudes, jargon, slang, etc. – that peak rapidly and then quickly die out. We’re all familiar with these: a cultural item – consider those annoying rubber wrist bands from a couple of years ago – suddenly shoots up in prevalence, appearing everywhere at once, and then suddenly pops back out of sight. So, fads by definition are trends that end quickly, but speed of uptake is obviously a key signal. In this post we’ll look at two different accounts of the relation of speed to ends and how our evaluation of fads differs from our evaluation of other trends. Finally, we’ll try to figure out what is “symbolically” at stake in the distinction between a trend and a fad and how this makes the latter automatically less worth joining.

First of all, Luis Bettencourt presents a model of the trend life cycle in which agents use a trend's relative speed of adoption as an indicator of viability and value. The faster the speed of adoption relative to other competing items, the more attractive an item is. When the speed slows, as it inevitably will given a finite population, agents will begin to abandon the trend provided the speed falls below an individually determined critical level. In this model, speed is a positive factor as it indicates the potential value, a form of “social proof,” of a trend. But it also inevitably brings about the trend’s end. Because of the way the dynamics are structured, trends that move fast, will move even faster, quickly bringing about their own crash. These are clearly fads.

Contrary to Bettencourt’s model, Jonah Berger and Gael Le Mens claim [pdf] that a high perceived rate of adoption of an “identity relevant” cultural item decreases our likelihood of adopting or consuming it. Identity relevant cultural items are those that can be thought of or used by potential consumers as a means of communicating some desirable or esteem-generating information about who they are or – more likely – who they’d like others to think they are; identity relevant cultural items are conventionally meaningful signals of group-specific tastes, social affiliation, status, etc. If adoption is too fast, the item might be a fad, which is a bad investment in the social identity stakes because, by definition, they can’t sustain popularity.

An important difference between the two views is obviously the latter’s focus on “identity.” In Bettencourt’s model, the faster the better for all items; it’s the fact that velocity is unsustainable in finite populations that leads to collapse. In Berger and Le Mens’s view, beyond a certain velocity threshold, the faster the worse, at least for publicly visible cultural items with identity implications. They argue that it’s our concern for the “symbolic value” cultural items may have for the development and maintenance of social identity that explains the desire to avoid fads. High velocity may decrease the attractiveness an item might have because it could be just a “flash in the pan” or of fleeting popularity. In other words, faddish cultural items are a bad investment as they don’t maintain symbolic value. So, in both models, speed is information, but in Berger and Le Mens’s, identity concerns decrease the item’s attractiveness (i.e. likelihood we’ll partake) as speed increases.

Their goal is to show that we judge potential faddishness by velocity and that we avoid fads because they're viewed negatively, which might be tied to the fear that they don't return symbolic value. Though they never really get into why we view fads negatively, I feel that we can say something stronger than that fads are unattractive because they’re bad social identity investments. Indeed, it seems that public association with a fad in an identity relevant domain may ultimately deliver disvalue as opposed to just decreased value. It not only won’t add to your social identity, in some situations it might actually damage your social identity. It’s often embarrassing or somehow dis-estimable to have been caught in a fad, to have publicly invested in or consumed some short-lived and now unpopular cultural item (if there are any pictures floating around the internet of you earnestly rocking a white Miami Vice blazer and woven loafers, you know what I mean).

So, I think that with a little reflection most will admit that getting publicly caught in a fad is somehow embarrassing or to be avoided, i.e. a disvalue. Can we articulate what it is about faddishness that actually bugs us? Why should the perception of an identity relevant cultural item’s faddishness make it less valuable or even potentially disvaluable? If it’s just that “flashes in the pan” – items that don’t sustain popularity/potential symbolic value – provide limited or no return to social identity on investment, why do we actually feel embarrassed about fad association as opposed to just annoyed at the wasted time? What is symbolically at stake in the distinction between fads and trends?

Autonomy and Constancy

Our social identity is impacted – often mediated – by the cultural items we consume or otherwise associate ourselves with. Cultural items have symbolic value related to what consumption of the item conventionally means or communicates. For example, the shoes you buy do more than just protect your feet or give you extra purchase on slippery sidewalks. They also communicate something about you, your tastes and often your position in a socio-cultural taxonomy (you’re a hipster in Converse or narrow slip-ons; an urban kid in puffy, tricked out Nikes; etc.). They’re public signals with conventional cultural connotations that you can exploit to manage your social identity. It’s these conventional connotations, the realm of taste and tastes, that confer the symbolic value. We manage our social identities partially by managing the set of cultural items we publicly associate ourselves with.

But clearly it’s not the cultural item alone that confers the symbolic value we exploit in managing perceptions of social identity. Perceptions of our motives and the longer term regularity of our social identity also impact our ability to wring symbolic value from a cultural item. For example, I had a friend who rapidly cycled through “personas,” going from punk to skinhead to b-boy to truck driver. With each new identity came a boatload of highly appropriate, conventionally meaningful gear, slang and comportment. But no matter how cool or dead-on they were for the current identity, he could never really wring any value from them. In fact given the suddenness of each of his transformations, they just seemed like cynical accoutrements, rendering symbolic disvalue and actually damaging his social identity to those he presumably most esteemed. The more he desperately tried to construct a "cool" or valuable social identity, the less likely it became that he actually could. Clearly, the items' perceived appropriateness or continuity with past social identity is important. More generally, perceptions of the motives for associating oneself with cultural items effect the items' symbolic value (and possibly one's larger social identity). If it seems out of character or if the social identity motive is too obvious, we likely won’t be able to garner any actual symbolic value from them no matter what they are.

So, in managing our social identities, we also have to manage perceptions of our management; there are “perceived motive” conditions on the symbolic value we get from some cultural item in an identity relevant domain. More specifically, there are autonomy and constancy conditions, which, if not met, decrease or possibly reverse the symbolic value available from a cultural item. Fads, pretty much by their nature, trash these conditions. Before saying why, let me look a little more closely at the perceived motive conditions, which must be met for an item to confer symbolic value.

Sometimes people try too hard to impress. They obviously speak, dress or comport themselves in what they assume relevant others consider the “cool” way. People like this are often called posers (or poseurs... but when I spell it that way I feel like one). They could be doing the same thing everybody else is doing, but their identity relevant moves are perceived as desperate, “inauthentic,” or even cynical.

As Jon Elster says, “nothing is so unimpressive as behavior designed to impress.” I think this “general axiom” extends to our public association with cultural items as well. Of course, we all recognize that our tastes are shaped by our in-group and larger culture. If not for the conventions and norms against which we evaluate cultural items and social identity, “symbolic value” really wouldn’t exist. But when identity relevant choices appear unduly concerned with others’ perceptions they start to lose value and may even damage social identity. When it appears that you’re dressing or talking a certain way solely out of concern for others’ perceptions of you, you may seem, for example, pretentious, conformist or cynical as opposed to cool. In short, in the West at least, we disvalue identity relevant moves that appear to be completely externally or cynically motivated, while we greatly value those that appear to be internally, “authentically” or autonomously motivated.

When it comes to your social identity you don’t want to appear that you’re trying on personas. This is closely related to autonomy, but conceptually distinct. You can switch your style daily, trying to align yourself with various cultural groups, in which case you’re a poser or social butterfly. But you can also switch daily just because you’re an odd loner, a kook. In this case, you’ve high autonomy, but low constancy. They’re distinct, but either way fickleness when it comes to identity is frowned upon. If you give the impression that you’re actively searching for an identity, it’s unsettling: it feels stagey, shifty and possibly cynical.

Though research suggests that our perceptions and expectations of singular self-identity are sort of illusory – more a product of self-narration and fundamental attribution error than some continuous, thing-like Self – we clearly assume and value the idea of constancy, consistency and continuity when it comes to social identity. Pomo identity theories notwithstanding, fickleness, or publicly searching for some sort of social identity, is often interpreted as dishonest, “inauthentic,” self-defeating, cynical and sometimes even pathological. It’s probably a cultural artifact of the West, but we really like to think of ourselves and others as Selves in the ideal sense. Acting otherwise can turn social identity moves on their head, making them seem like a sham and destroying any potential value.

Why We Avoid Fads

We are socially incentivized to manage our social identity management. We seek to construct social identities that are meaningful for and align with existing socio-cultural groups. But in order to create a valuable social identity, we have to construct it in a way that gives the impression that our choices are the autonomous decisions of a constant, stable, already fully constituted social identity. Failing to give the impression of autonomy and constancy in our decisions decreases the “value” of social identity relevant moves.

Fads are culturally salient; their sudden appearance everywhere gives them high visibility and holds our focus. Because of our (probably cultural) bias toward autonomy and constancy – that is, for evaluating identity moves against the ideal of individual, reflective choice by a singular Self – we interpret these waves of uptake and failure in individual terms. After a fad crashes, our bias pushes us to interpret it negatively as a case of social influence over autonomous decision, as a case of herd mentality. The speed of the descent we interpret as social proof of the emptiness or valuelessness of the fad. But the speed of the ascent we interpret as the dis-estimable actions of non-autonomous, inconstant conformists: it’s the result of unreflective "bandwagoneers," people with no strong Selves embarrassingly misled by social influence.

Provided you want to optimize the potential symbolic value from a cultural item, you had better take into consideration whether or not it’s a fad. But, that’s not the only reason you should avoid them. Publicly associating with fads may damage your larger social identity. That is, association with a fad is interpreted as a failure of the autonomy and constancy conditions and specific failures of general conditions reflect on all of your identity relevant decisions. Publicly joining a fad changes the way others interpret your motives for all social identity relevant decisions; your motives become suspect not just in this case, but to some limited degree in all past and future cases. Provided information is publicly available, association with fads incrementally whittles away at the perceived "authenticity," "autonomy" and, effectively, "value" of your social identity. So, it’s a wise strategy to wait and see. Failing that, use the information at hand to judge the probability that some cultural item is a fad. Keep your eye on the speedometer; velocity of uptake is the most salient indicator of a fad pre-crash.


Amos said...

I can't wait until your archives are translated into French

Shannon Bain said...

me either... then maybe somebody would read them.

Amos said...

As you have presented them, Bettencourt appears to be speaking of adoption, while Berger & Le Mens seem to speak of the perception of adoption, which is a meta-analysis.

There may be agents sitting anywhere along the curve: someone at the very front of the curve will have no idea how big the fad will get or how long it will last - maybe that is how this type of agent likes to live; another agent in the middle of the curve sees and is attracted to the amplitude, but has little-to-no concept of its period and rate of decline - for these you could say the existent popularity is the attraction; an actor at the end of the curve may see it has a declining amplitude, but still jump on for personal ironic or cynical reasons.

At a higher level, you could say that the existence of the curve, and its rate of growth, give it a de facto value, regardless of who sits where.

The agent-based analysis also brings up issues of the spread of information. An actor who jumps on at the end of the curve may not have enough of a birds-eye view (i.e., the public availability of information you mention at the end of the post) of the situation to know that this fad is on the decline. And perhaps within that actor's microcosm, they think they are at the front, not the back, like wave passing through a hole in a wall and starting anew. It's all relative, dude.

So it seems that an analysis of agent-based motivations and perceptions, which your post is truly about, requires seperate tools than the overarching analysis of the epidemiological spread of information.

But anyway, there's no need to justify your refusal to accept anything but disposable plastic bags at the supermarket.

Shannon Bain said...

Thanks for the comments. To clarify, Bettencourt is just building an agent based model. Berger and Le Mens are providing an interpretation of an econometric analysis of baby-name adoption rates.

So, yes, Bettencourt is clearly speaking of adoption, but the information pertinent to the agent's adoption decision is relative speed of uptake. So adoption depends on perception of velocity. Beger and Le Mens suggest that velocity is an (probabilistic) indicator of "period," i.e. the probability that it will be a fad (short period) or a lasting trend (longer period). (I would argue, incidentally, that irony isn't cyclic or periodic. Check out my post on cultural irony.)

As to amplitude, in Bettencourt's model, it seems that fads, as high velocity and therefore high attraction trends, will automatically be high amplitude. However, on Berger and Le Mens's account, fads can be of relatively low amplitude, as the initial perception of velocity actually suppresses incipient adoption. So, on both accounts, velocity is both an indicator and creator of fads, but only on the former does it carry information about amplitude. I was trying to theorize about why it would suppress.

You're right to point out that this is about information "conveyed" in agent's actions. But in addition it's about agent's attitudes or dispositions given the information conveyed. Berger and Le Mens contend, rightly I think, that agents have an aversion to association with fads. Again, I was trying to give an account of the aversion by considering how the in-group, through which you form your social identity, evaluates your use of identity relevant cultural items. It's not just about the item; it's also about perceptions of your use of the item in building identity. I take these latter to be capable of damaging your identity if not appropriately handled.

They don't really have many plastic bags here in Switzerland. Mostly expensive paper bags... greenest country in the world after all.

Shannon Bain said...

One last thing. Regarding the distinction between adoption and perception of adoption, I'm the one that discusses the perception of adoption – i.e. tries to give reasons for the aversion to fad association – while Berger and Le Mens discuss adoption – whether or not to associate with some item given that it may be a fad. They don't really analyze the aversion, but I try to.