I'll be pretty disappointed if our community accepts the idea that humiliating other people has less to do with status than comparing Pokemon collections. Which situation makes you more conscious of status: when your Pokemon collection is smaller than Bob's, or when Bob beats you up and takes your girlfriend? To really feel the concept, you have to be close to the monkey life.

close to the monkey life

That sounds like an allusion to dominance hierarchy theory, which my informal survey suggests is a muddle. Do you have pointers to solid, recent research on dominance hierarchy theory that could plausibly apply to humans?

People do sometimes react strongly to things we think weird, like not having the bigger Pokemon collection.

5RobinZ10y This is true, but status still isn't about who can beat up who.

The many faces of status

by Morendil 10y15th Apr 20106 min read109 comments


The term "status" gets used on LessWrong a lot. Google finds 316 instances; the aggregate total for the phrases "low status" and "high status" (which suggest more precision than "status" by itself) is 170. By way of comparison, "many worlds", an important topic here, yields 164 instances.

We find the term used as an explanation, for instance, "to give offense is to imply that a person or group has or should have low status". In this community I would expect that a term used often, with authoritative connotations, and offered as an explanation could be tabooed readily, for instance when someone confused by this or that use asks for clarification: previous discussions of "high status" or "low status" behaviours seemed to flounder in the particular way that definitional arguments often do.

Somewhat to my surprise, there turned out not to be a commonly understood way of tabooing "status". Lacking a satisfactory unpacking of the "status" terms and how they should control anticipation, I decided to explore the topic on my own, and my intention here is to report back and provide a basis for further discussion.

The "Status" chapter of Keith Johnstone's 1979 book "Impro", previously discussed here and on OB, is often cited as a reference on the topic (follow this link for an excerpt); I'll refer to it throughout as simply "Johnstone". Also, I plan to entirely avoid the related but distinct concept of "signaling" in this post, reserving it for later examination.

Dominance hierarchies

My initial impression was that "status" had some relation to the theory of dominance hierarchies. Section 3 of Johnstone starts with:

Social animals have inbuilt rules which prevent them killing each other for food, mates, and so on.  Such animals confront each other, and often fight, until a hierarchy is established, after which there is no fighting unless an attempt is made to change the ‘pecking order’. This system is found in animals as diverse as humans, chickens, and woodlice.

This reinforced an impression I had previously acquired: that the term "alpha male", often used in certain circles synonymously with "high status male", indicated an explicit link between the theoretical underpinnings of the term "status" and some sort of dominance theory.

However, substantiating this link turned out a more frustrating task than I had expected. For instance, I looked for primary sources I could turn to for a formal theoretical explanation of what explanatory work the term "alpha male" is supposed to carry out.

It seems that the term was originally coined by David Mech, who studied wolf packs in the 70's. Interestingly, Mech himself now claims the term was misunderstood and used improperly. Here is what David Mech says in a recent (2000) article:

The way in which alpha status has been viewed historically can be seen in studies in which an attempt is made to distinguish future alphas in litters of captive wolf pups [...] This view implies that rank is innate or formed early, and that some wolves are destined to rule the pack, while others are not.

Contrary to this view, I propose that all young wolves are potential breeders and that when they do breed they automatically become alphas (Mech 1970). [...] Thus, calling a wolf an alpha is usually no more appropriate than referring to a human parent or a doe deer as an alpha. Any parent is dominant to its young offspring, so "alpha" adds no information.

An informal survey of other literature suggests that "alpha male", referring specifically to the pack behaviour disowned by Mech, entered the popular vocabulary by way of dog trainer lore. My personal hunch is that it became entrenched thereafter because it had both a "sciencey" sound, and the appropriate connotations for people who adhered to certain views on gender relationships.

Stepping back to look at dominance theory as a whole, I found that they are not without problems. Pecking order may apply to chickens, but primates vary widely in social organization, lending little support to the thesis that dominance displays, dominance-submission behaviours and so on are as universal as Johnstone suggests and can therefore be thought to shed much light on the complex social organization of humans.

An often discussed example is the Bonobo chimpanzee, where females are dominant over males, and do not establish a dominance hierarchy among themselves, whereas males do; where the behaviours that tend to mediate social stratification is reconciliation rather than conflict, something that is also observed in other animal species, contrary to the prevailing view of dominance hierarchies.

This informal survey was interesting and turned up many surprises, but mostly it convinced me that dominance hierarchies were not a fruitful line of research if I was after a crisp meaning of "status" terms and explanations: either "status" was itself a muddle, or I needed to look for its underpinnings in other disciplines.


Social stratification

Early on in Johnstone there is an interesting discussion of status by way of his recollection of three very different school teachers. At various other points in the chapter he also refers to the stratification of human societies specifically, for instance when he discusses the master-servant relationship.

The teacher example was particularly interesting for me, because one of the uses I might have for status hypotheses is in investigating the Hansonian thesis "Schools aren't about education but about status", and what can possibly be done about that. But to think clearly about such issues one must, in the first place, clarify how the hypothesis "X is about status" controls anticipation about X!

I came across Max Weber (who I must say I hadn't heard of previously), described as one of the founders of modern sociology; and Weber's "three component theory of social stratification", which helped me quite a bit in making sense of some claims about status.

What I got from the Wikipedia summary is that Weber identifies three major dimensions of social stratification:

  • class or wealth status, that is, a person's economic situation
  • power status, or a person's ability to achieve their goals in the face of other's opposition
  • prestige status, or how well a person is regarded by others

This list is interesting because of its predictive power: for instance, class and wealth tend to be properties of an individual that change slowly over time, and so when Johnstone refers to ways of elevating one's status within the short time span of a social interaction, we can predict that he isn't talking about class or wealth status.

Power status is more subject to sudden changes, but not usually as a result of informal social interactions: again, power status cannot be what is referred to in the phrase "high status behaviours". Power is very often positional, for instance getting elected President of a powerful country brings a lot of power suddenly, but requires vetting by an elaborate ritual. (Class status can often go hand in hand with power status, but that is not necessarily or systematically the case.)

Prestige status can be expected to depend on both long-term and short-term characteristics. Certain professions are seen as inherently prestigious, often independently of wealth: firemen, for instance. But within a given social stratum, defined by class and power, individuals can acquire prestige through their actions.This is applicable for wide ranges of group sizes. Scientists acquire prestige by working on important topics and publishing important results. Participants in an online community acquire prestige by posting influential articles which shape subsequent discussion, and so on.

But, while it struck me as conceivable to unpack terms like "high status behaviours" as referring to such changes in prestige status, it didn't seem entirely satisfactory. So I kept looking for clues.


Self-esteem and the seesaw

Johnstone refers to status "the see-saw": he sees status transactions as a zero-sum game. To increase your status, he says, is necessarily to lower that of your interlocutor.

This seems at odds with seeing most references to status as meaning "prestige status", since you can acquire prestige without necessarily lower someone else's; also, you can acquire prestige without entering into an interactive social situation. (Think of how a mountaineer's prestige can rise upon the news that they have reached some difficult summit, ahead of their coming back to enjoy the attention.)

However, most of what Johnstone discusses seemed to make sense to me if analyzed instead as self-esteem transactions: interactive behaviour which raises or lowers another's self-esteem or yours.

There is lots of relevant theory to turn to. Some old and possibly discredited - I'm thinking here of "transactional analysis" which I came across years and years ago, which had the interesting concept of a "stroke", a behaviour whereby one raises another's self-esteem; this could also be relevant to analyzing the PUA theory of "negging". (Fun fact: TA is also the origin of the phrase "warm fuzzies".) Some newer and perhaps more solidly based on ev-psych, such as the recently mentioned sociometer theory.

Self-esteem is at any rate an important idea, whether or not we are clear on the underlying causal mechanisms. John Rawls notes that self-esteem is among the "primary social goods" (defined as "the things it is rational to want, whatever else you want", in other words the most widely applicable instrumental values that can help further a wide range of terminal values). It is very difficult to be luminous, to collaborate effectively or to conquer akrasia without some explicit attention to self-esteem.

So here, perhaps, is a fourth status component: the more temporary and more local "self-esteem status".


Positive sum self-esteem transactions?

Where I part company with Johnstone is in seeing self-esteem transactions as a purely zero-sum game. And in fact his early discussion of the three teachers contradicts his own "see-saw" image, painting instead a quite different picture of "status".

He describes one of the teachers as a "low status player", one who couldn't keep discipline, twitched, went red at the slightest provocation: in other words, one with generally low self-esteem. The second he describes as a "compulsive high status player": he terrorized students, "stabbing people with his eyes", walked "with fixity of purpose". In my terms, this would be someone whose behaviours communicated low regard for others' self-esteem, but not necessarily high self-esteem. The third teacher he describes as "a status expert":

Much loved, never punished but kept excellent discipline, while remaining very human. He would joke with us, and then impose a mysterious stillness. In the street he looked upright, but relaxed, and he smiled easily.

To me, this looks like the description of someone with high self-esteem generally, who is able to temporarily affect his own and others' self-esteem, lowering (to establish authority) or raising (to encourage participation) as appropriate. When done expertly, this isn't manipulative, but rather a game of trust and rapport that people play in all social situations where safety and intimacy allow, and it feels like a positive sum game.

(These transactions, BTW, can be mediated even by relatively low-bandwidth interactions, such as text conversations. I find it fascinating how people can make each other feel various emotions just with words: anger, shame, pride. A forum such as Less Wrong isn't just a place for debate and argument, it is also very much a locus of social interaction. Keeping that in mind is important.)

Detailed analysis of how these transactions work, distilled into practical advice that people can use in everyday settings, is a worthwhile goal, and one that would also advance the cause of effective collaboration among people dedicated to thinking more clearly about the world they inhabit.

Let the discussion stick to that spirit.