Friday, 29 August 2008

The Bottom Line is I Hate Jargon: why hated language is inescapable

Everybody says they hate jargon, but everybody uses it. How can jargon be so hated yet so prevalent? It’s clearly more than just a personal shortcoming of your annoying co-workers. It’s more likely that jargon is a symptom of sociability and thus we’re all susceptible to it. Looking closely, there seem to be at least three types of irritating jargon (each with its own jargon-like name!):

Fluffy Neologisms (FuNs): Often a “verbized” noun or a verb phrase shortened to one word, e.g. conversate for have a conversation. Frequently an active sounding, metaphorical stock phrase for something simple, e.g. circle back for get back to you and take it offline for talk after the meeting. FuNs are most often emergent (bottom-up).

Pernicious Euphemisms (PEus): These are classic business speak, e.g. calling criticisms opportunities. The purpose is to mask some harsh reality, neutralize something negative or insinuate desired behavior (e.g. employees must be passionate – i.e. work-obsessed, myopic, ass-kissers – as opposed to just good at their jobs). PEus tend to be institutionally imposed (top-down).

Exclusionary Technicalities (ETs): These are jargon in the traditional sense of “words or expressions that are used by a particular profession or group and are difficult for others to understand”, e.g. AJAX interface or MVC Architecture. They’re domain specific “technical” terms intended to ease in-group communication. ETs can be either institutionally imposed (by e.g. standards bodies) or emergent (from e.g. open source developers).

We tend not to think of legitimate uses of ETs as jargon in the pejorative sense (unless you are particularly insecure). After all, these words have a technical rationale. What gets us miffed is inappropriate use by the tech guy in an effort to maintain his wizard status. So, for ETs it’s intention that matters. A lame, illegitimate use can turn an ET into a PEu and open the user up to justified criticism. Indeed, many words that end up as PEus probably started out as ETs in management science.

FuNs, though, seem inexcusable to most people. They probably stem from middle-management aping the “management science” consultants they caught the PEus from. And that’s part of my point: jargon is catching. It’s not simply some aggravating sign of your boss’s personality disorder. Rather, it’s a symptom of sociability which we all display; it's an equilibrium in a social market like a fad or a trend. Of course, there is the pretension to hate, but generally before the pretension there was the desire to communicate something in a codified way.

The net net on jargon...

Typically, we’re aggravated by jargon in one of the following three ways.

1. We consider it pretentious when the user is too obviously attempting to wring cultural or social profit from what should be a purely instrumental linguistic exchange.

2. We’re exasperated when some overused phrase, which has been nearly drained of significance by loose use, is unreflectively trotted out. This sort of hackneyed jargon is often used more as a combination filler/badge than as a thoughtful addition to the discourse, serving no purpose in the conversation other than a noise our type tends to make in situations like this one.

3. Worst of all, we’re infuriated when someone mixes 1) and 2), using hackneyed jargon pretentiously.

But if you look at these from the other direction – from the jargon user’s perspective – you can see how the user could derive value from using jargon. After all, people don’t use jargon in order to be negatively viewed by others; presented with socially informed options, they have made a choice, albeit a non-deliberative one. When we have questions about a choice, look to the expected value of the options for some answers.

Let’s start with the most straightforward example: legitimate uses of Exclusionary Technicalities (ETs). You can see how these are valuable: they are agreed upon domain specific names for objects, phenomena, etc. If we want to speak efficiently and non-circuitously when discussing these things with our peers, we should use the jargon. The more people who use the term for the intended thing, the more valuable the term becomes in the field and the more value it holds for a potential adopter. After a certain percentage of your peers start using the jargon it behooves you to climb aboard. We call this a network effect: the more folks using the jargon, the more valuable it becomes. Network effects can lead to cascades (where everyone sees the value of adopting the jargon as greater than not adopting) and suddenly the jargon is everywhere.

People legitimately using ETs tend to value communication; the jargon’s value to them is in speaking the language their peers speak. The “value function” is stable, producing a value that rises rapidly to a plateau and maintains high value for a long time provided there are no relevant shocks (say, new standards rendering the old ones obsolete). We call this a monotonic externality. The value function tends to go in one direction, up, until the plateau. It doesn’t suddenly reverse with changes in use of the term.

Illegitimate uses of ETs, however, are a slightly different story. For example, I hear of a technology that’s getting a lot of buzz in the tech blogs and that I’m sure all of my clients will want soon. I drop the new jargon in meetings, making myself more attractive to the client. Of course, all other strategists are doing the same thing. The ET, which solidified it’s value in the tech world enough to be noticed in our non-tech world, is clearly of value to a lot of smart people. So, we assume its value without doing any sort of true evaluation. Many other people do the same thing. Next thing you know, everybody’s talking about and asking for AJAX interfaces, or MVC Platforms, yet very few know just what these are. These ETs have become jargon in the bad sense. Now we’re in an Informational Cascade: given a certain number of visible and presumptively knowledgeable adopters, people stop evaluating and go straight to adopting. The more the jargon’s exposed by adopters, the faster it’s adopted by others.

However, informational cascades are very unstable. Once someone actually does stop and try to figure out what’s being talked about, the cascade can be very quickly shattered. Ideas that spread like wildfire can die out just as fast. The jargon then becomes an embarrassment... what was the big freakin’ deal about “Web 2.0”? Then a backlash can set in.

PEus and FuNs are the prime examples of this sort of success followed by backlash. Despite their differences, they tend to follow the same general cycle: acceptance and use followed by derision and annoyance.

This “value function” that increases rapidly creating a cascade and then crashes just as rapidly after a certain point seems to be what’s called a non-monotonic externality. That is, the value of adopting becomes negative after a certain level of adoption within your group; after a certain level of saturation it becomes uncool, embarrassing or simply over-inflated like a linguistic bubble in a socio-cultural market. A value curve like this looks just like a fad or fashion curve.

In this case, the value function is pretty interesting. There’s an element of “network externality” in that jargon serves as shorthand in recurring situations and it’s assortative, defining your in-group. For these reasons, it behooves you to adopt if others in your group are. But then there’s also an element of informational cascade in that we value adoption simply because others are doing it. It’s about unreflective conformity and social learning. So it’s coordinating to a large extent. Still the curve suggests that difference is significant. That is, the value function delivers disvalue given a certain saturation. But our experience with offices is that negative feelings don’t set in until much later than most fads. Often, an office has to be lousy with a particular bit of jargon before we start to hate it.

Everyone’s “value function” is different, of course. Some get annoyed very easily (“misfits”) and others never really get annoyed (“team players”). Also, whether we evaluate jargon negatively or not is very contextual. I could listen two techies spewing jargon at each other all day and not care, but if one of them scares a client with a string of acronyms, I could really get annoyed.

Basically, the big moral is that we all tend to use jargon and it’s not a personal shortcoming. Rather it’s the result of very common social market conditions. Social markets are noted for each actor’s actions being partly determined by the actions of others. Dynamics like this lead to equilibria, but because of things like the fragility of informational cascades and non-monotonic externalities, they are fragile equilibria. We start to disvalue jargon that was once valued (even if non-deliberatively) for, among other reasons, the desire to feel distinct. We don’t like to feel conformist, while at the same time we clearly value a certain amount of conformity (viz. network effects and informational cascades). That’s just the nature of a non-monotonic value curve; the fact that something’s being used by our peers is reason enough to start doing it, but after a certain level of saturation we start to disvalue what we once valued (we can think of slang in similar terms, though the value curve is usually much shorter, thus it’s much more fragile or fickle). So, the conformity arises, but is rebelled against after a certain level of saturation.


joyhoward said...

geez, I hope no one refers to me as a passionate, team-player tomorrow. if so, I will try not to point out the obvious, uncomfortable non-monotonic externality. . . which I suppose would be an ET for your blog readers

Kaveeta Sehjal said...

do you have any references for this blog?
if so, please can put them in a comment


Claire Martin said...

Loved this explanation! I am a teacher, and the jargon is epic. Sometimes I interrupt coworkers who say words like "protocols" and yell words like "rules" at them. I will probably never be principal.