The amateur Anthropologist
Hey I said these were randome thoughts. My reall aim is to prove that there is someone who can't concentrate long enough on a topic to have a blog.
20 years ago I had an Idea. Maybe someone who knows something about the field can tell me what is wrong with it in 20 seconds (including maybe someone else had the idea 40 years ago).
This thought was stimulated by reading Structural Anthropology a collection of essays by Claude Levi-Strauss. There are two questions. One is why are some cultures monogynous and others polygynous ? The other is why do the Bororo divide their tiny villages into 3 endogamous clans ?
OK first question. Why in some cultures men can marry more than one woman and in others only one ? One possbile explanation is polygyny occurs when the gender ratio is many women for each man. This can happen if lots of men get killed by other men. So women share husbands or go single wasting their uteruses (the Moll Flanders problem described by Daniel Defoe some time ago). Could be the explanation, but I would like to talk about another. Levi Strauss was very interested in a very simple mathematical model which pointed out that hunter gatherers typically live in tiny groups (have too to avoid killing off all the game within walking distance). Someone else (really some two else) calculated that these groups were about as small as could be sustained given risk that a generation would be all male or all female and thus the last (he didn't explain this model very clearly and I didn't look it up). OK see how much worse this problem is if monogynous. If people live in small groups and are mostly endogamous (must have some flow with other villages/bands to avoide inbreeding but I assume this is pretty low). If each man is allowed to get only one woman pregnant, the number of woman who reproduce each generation is the lesser of the number of woman and the number of men. If each man is allowed to get as many women pregnant as are available then the number of women who reproduce each generation is the number of women. Polygyny might be required in people who live in small mostly endogamous villages to deal with random fluctuations in the sex ratio.
Now What's the problem ? Well in farming, herding or industrialised societies there is the problem that women who share a husband are poor (leave out issue of single moms in cultures which are trying to be monogynous but don't do a very good job of it). This is not the problem in hunter-gatherer cultures where food is shared within the village. The economic unit is not the family but the village (has to be that way given risky returns to hunting not to mention hunting often works best with village size group cooperating).
So is there a disadvantage of polygyny ? I think there is (not that I have any personal experience (REALLY I swear dear my interest is purely theoretical really -- no field trials)). I think it makes the rival suitors problem worse. Think of this little village with two guys chasing after the same gal. That's tension. Monogyny puts a limit on the allowed licit romantic ambitions of each man which reduces the conflict over the babes. We can see in other primates this conflict is very very intense. OK for them as they aren't armed. I think humans have this problem that we have developed weapons which can make quarrels lethal so we men have to stop fighting over the women. Clearly this problem is not easy not solved etc. Now polygyny might (or might not) reduce conflict between women over the men (I also have no experience with polygyny from the wive's point of view). However, women are much less violent than men (a cultural universal I think except where violence levels are so low that it is hard to tell).
Finally I am getting to Levi-Strauss' main interest -- Kinship rules. The observation is that in hunter gatherer populations there are often very strict narrow restrictive rules on who can marry who. There are many theories of why. Levi Strauss has an attractive theory that the idea is to make sure that everyone is related to everyone else in the village so kinship bonds hold the village together. Very nice theory but what about the Bororo ? They have rules restricting people to marry withing their "clan" of which there are 3 in each village. This sure doesn't fit L-S' theory since it divides each village into 3 kinship groups instead of uniting it.
What does he write about them ? He writes that the Bororo made a mistake. That they don't understand how kinship rules should work (and stubbornly refused to listen to his explanation I suppose). That, as a result, their culture is doomed (don't believe me look it up). Hmmm so this is the methodology of structural anthropology according the L-S ? The argument is we find patterns of culture because of cultural selection -- some combinations lead to stability and are selected others to instability and are not selected. Now here is a culture which should not exist according to the theory. Conclusion -- it won't exist for long. Notice how a theory of selection in which observed cultures might be un fit (no problem) is a theory without any predictive content.
OK how about my story. According to my story the point of these complicated rules of who can marry who is to avoid conflict between two men who each want to marry the same woman. Thus the point of the rules is that they are rules. It doesn't matter much if they tell this poor woman miss Smith that she is going to have to be mrs Jones or mrs Miller. The point is they leave her no choice so she can't break any hearts.
Personally I would prefer a doomed unstable culture, but I think the argument makes cultural evolutionary sense. If it's all decided in advance as soon as people are born there is no point in fighting over it.
Now back to the Bororo. The 3 clans make no trouble for my story at all. The point is to have rules. The exact rules don't matter so much. A culture which divides villages into arbitrary groups and forces them to be endogamous is like a culture which divides them into arbitrary groups and forces them to be exogamous. In either case men can't fight each other for wives unless one sets himself against culture and tradition.
One final point. I assume that the rules are not binding to the point that women can't marry if they have no allowed husband. That is, I assume they are really rules for deciding which of more than one candidate husband is chosen. If women are forced to be single because of the kinship rules, my theory is dead wrong. That's a very weak testable prediction. I also predict stricter rules of kinship if polygynous than if monogamous. A vague prediction.
Sunday, October 06, 2002
10 seconds ago I had an idea. Maybe in 10 seconds I will see what is wrong with it, but right now I like it a lot.
problem: when we need a killer t-cell response we only make antibodies which don't do the job.
Proposed solution: Make a peptide which has the fab part of the antibody fused to some peptide which just drives our killer cells nuts with fury. Inject it. the peptide goes and sticks to the target (no big deal we already have antibodies doing it). There it sits waving it's tail end which is like a red flag to a bull. The killer cells come and kill the target.
Why not ? Makes sense to me.
Now to work we need activated (or memory) killer cells around. The red flag will not be presented on an antigen presenting cell. That means must pre-stimulate with some known peptide to get memory killers of right type.
Problem includes tumours and HIV infected cells so two really big problems.
OK now more than 10 seconds have passed and I see what's wrong with the idea (was a very exciting minute anyway). HLA restricted. Ugh. seems would at least need a different construct for every HLA type. So the known peptide could be HLA type 1 presented by HLA type 2.
Back to drawing board. First could there be a more general red flag ? Like a monoclonal which sticks to CD8 and constant region of t-cell receptor ?
Like some peptide which happens not to give a killer response as HLA restricted as usual ?
Or maybe the plan is a huge massive semi mechanised hybrid construction project. We want Fab end sticks to HIV infected cells (or lung cancer cells) and red flag end looks like HLA type 1 presented by HLA type 2 through N (N is number of HLA types) or HLA type 2 presented by HLA type 1.
I'm thinking as I type so this is getting long.
New idea
There are monoclonals which stick to something we don't like and want to kill (cancer cell or hiv infected cell).
We each have memory type killer cells waiting to kill something (different for each of us which is the problem).
Now what about helping me Robert Waldmann (so far I don't have any cancer I know of and don't have HIV but I am being honest about how time consuming the project is becoming) Can we make a red flag which activates a memory type killer cell which I already have ? Well if we are making a peptide, probably best approach is not to try to make peptide presented by HLA but rather an monoclonal which sticks to and activates a memory type killer already there.
Question: Are there known monoclonals stick to t-cell receptor andactivate memory type killer cells. If not not so helpful to go on. It seems to me it should be possible.
OK new peptide to kill target cell is monoclonal sticks to target cell linked to monoclonal sticks to AND stimulates t-cell receptor of a memory type killer cell.
Notice the change. As someone who never managed to construct a construct, I am reluctant to propose constructing a large number (number of targets
times number of HLA types). Now the link the monoclonal which sticks to target to the red flag for killer t-cells is a link to protiens. Indeed
they are antibodies so can make a hybrid by breaking S-S bonds and remaking them. That is current version of idea works without using restriction enzymes or ligase.
So for proof of principal the stick to target AB could be something which sticks to some other leukemia (easily accessible cells).
The red flag AB is make antibodies to t-cell receptor of memory t-cells (Isolated by FACS) and screen for agonist antibodies which drive them into a homicidal (sorry cytocidal) fury.
Link ab's inject and hope.
problem: when we need a killer t-cell response we only make antibodies which don't do the job.
Proposed solution: Make a peptide which has the fab part of the antibody fused to some peptide which just drives our killer cells nuts with fury. Inject it. the peptide goes and sticks to the target (no big deal we already have antibodies doing it). There it sits waving it's tail end which is like a red flag to a bull. The killer cells come and kill the target.
Why not ? Makes sense to me.
Now to work we need activated (or memory) killer cells around. The red flag will not be presented on an antigen presenting cell. That means must pre-stimulate with some known peptide to get memory killers of right type.
Problem includes tumours and HIV infected cells so two really big problems.
OK now more than 10 seconds have passed and I see what's wrong with the idea (was a very exciting minute anyway). HLA restricted. Ugh. seems would at least need a different construct for every HLA type. So the known peptide could be HLA type 1 presented by HLA type 2.
Back to drawing board. First could there be a more general red flag ? Like a monoclonal which sticks to CD8 and constant region of t-cell receptor ?
Like some peptide which happens not to give a killer response as HLA restricted as usual ?
Or maybe the plan is a huge massive semi mechanised hybrid construction project. We want Fab end sticks to HIV infected cells (or lung cancer cells) and red flag end looks like HLA type 1 presented by HLA type 2 through N (N is number of HLA types) or HLA type 2 presented by HLA type 1.
I'm thinking as I type so this is getting long.
New idea
There are monoclonals which stick to something we don't like and want to kill (cancer cell or hiv infected cell).
We each have memory type killer cells waiting to kill something (different for each of us which is the problem).
Now what about helping me Robert Waldmann (so far I don't have any cancer I know of and don't have HIV but I am being honest about how time consuming the project is becoming) Can we make a red flag which activates a memory type killer cell which I already have ? Well if we are making a peptide, probably best approach is not to try to make peptide presented by HLA but rather an monoclonal which sticks to and activates a memory type killer already there.
Question: Are there known monoclonals stick to t-cell receptor andactivate memory type killer cells. If not not so helpful to go on. It seems to me it should be possible.
OK new peptide to kill target cell is monoclonal sticks to target cell linked to monoclonal sticks to AND stimulates t-cell receptor of a memory type killer cell.
Notice the change. As someone who never managed to construct a construct, I am reluctant to propose constructing a large number (number of targets
times number of HLA types). Now the link the monoclonal which sticks to target to the red flag for killer t-cells is a link to protiens. Indeed
they are antibodies so can make a hybrid by breaking S-S bonds and remaking them. That is current version of idea works without using restriction enzymes or ligase.
So for proof of principal the stick to target AB could be something which sticks to some other leukemia (easily accessible cells).
The red flag AB is make antibodies to t-cell receptor of memory t-cells (Isolated by FACS) and screen for agonist antibodies which drive them into a homicidal (sorry cytocidal) fury.
Link ab's inject and hope.