The many faces of status

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.