There always could be. At what point do you believe Occam's Razor kind of kicks in?
When the coincidences or secondary effects are not just as likely given the evidence available.
it makes me wonder exactly what scientific research you do accept.
I've already said I accept scientific conclusions, even within the nonpop portions of EP. What I use to select whether I trust conclusions is proper methodology.
Here is my analysis:
For those who have not been following along, the study I'm looking at can be found at
this link. All quotes will come from there.
BRIEF STUDY SUMMARY:
The study describes that phenotypic body symmetry has been connected to better fitness/health components. There are also reported connections between symmetry and visual attractiveness (which makes sense based on its obvious phenotypic qualities). Separately, women detect scents variably over menstrual cycle, specifically ones related to sexuality (or at least those are the ones which have been tested). The authors attempt to replicate an earlier (and admittedly flawed) study of their own which indicated a correlation between scent attractiveness and symmetry in men.
For argument's sake previous research regarding symmetry will be assumed valid, except to note that the earlier scent study was not rigorous, which means that its results cannot be assumed valid, though it will lead to conduct of other studies (such as this one).
The concept of the authors is that a chemical product is associated with male body symmetry (both having been genetically driven together in some manner), and that females have adapted sensory processes to detect those particular chemicals, in order to select "good genes", just as they have selected visual processes to find "good genes" by finding visual symmetry more attractive. This is supposed to be shown more "convincingly" if women correlate them more highly during times of high fertility.
There are some positive elements to this research, and indeed I am interested in interplay of hormones, including secretion and detection. But I am going to focus on problems because this may be lengthy enough.
THEORETICAL
One has to remember that what they are suggesting is rather a large claim to make. They are not simply trying to locate chemicals that females may cue on during their menstrual cycle, and then find what that may be associated with, and then see what role that may play in survival (or even document its existence demographically and physiologically within men or women).
They are jumping through several assumptions in order to link women's attraction to some substance via scent and its correlation to symmetry, to argue this indicates some genetically driven adaptation with a positive reproductive purpose. Women are genetically keyed to detect X for specific associations.
This raises questions.
1) Symmetry (or its opposite: asymmetry) within any individual has both genetic and environmental factors. The authors do not reveal evidence suggesting that the genetic component is a predominant factor, but assert...
FA variation should reflect heritable fitness much of the time, often making FA an indicator of genetic quality and leading to selection for mate preferences based on FA or on traits such as scents that may covary with FA.
Should reflect, much of the time, and often making. This does not present a state of knowledge such that one can confidently suggest that symmetry would lead to selection for overt mate preference, based on secondary traits that "may covary". This is speculation, and assumption, without weeding out alternative possibilities of how one of those secondary traits might arise and stay within a population.
2) Female cues within menstrual cycles, do not necessarily have to be coded for selecting good genes. It would seem reasonable to postulate that females will experience different sensory experiences during different parts of their cycle. Hormones may make their processing extra sensitive, thus heightening all sensations, or making them selectively heightened. While those with "good genes" may be rewarded, the cues females find attractive may specifically be related to cues on current health or vigor, rather than "overall fitness" capability.
Indeed I am sort of baffled that the authors seem to operate under a concept of "good genes" when such a thing is itself relative to changing environments. If for some reason a person with asymmetric gene-disposition developed an advantageous mutation (lets say immunity to HIV) it would be self-defeating to prefer symmetry gene-disposition. They would become "bad genes" very quickly.
Given the back and forth nature of life and environments it might make more sense that females cue to common traits regarding general present health which would be exhibited by both symmetric and asymmetric men relatively equally in similar health conditions. Thus those surviving well at anytime, and in any environment, would stand out, rather than simple symmetry which may have an advantage under most conditions only at this time.
3) Female cues do not necessarily have to be genetically fixed. I did not see any discussion of research investigating how female scent prefs develop. It could be that females "learn" to associate smells (chemicals) with other traits that they find positive. In this specific case if symmetric individuals put out consistently more or unique pheromones, females may over time create that association, such that it is imprinted as a selective factor. Organic based pref may have nothing to do with evolutionary/genetic hardwiring of selection factors.
4) While this next point could equally be in the results/analysis problems section, I think it also belongs here, because it represents the deductive theoretical problem that the authors exhibit. They state at one point...
In sum, we know of no explanation of our results other than that normally ovulating women use a chemical(s) in men's sweat or skin as a basis for discriminating men who have and have not experienced perturbations that generate FA
They start with this position, make assumptions such that a certain correlation would act as an indicator if the assumption were true, and after getting that correlation state they can't think of any other explanation. If you think you've heard that before, you probably have. It was ID theorists using the same argument to invoke design. While these authors may at least be attributing purpose to an unknown process rather than an unknown creator, the argument is the same... incredulity.
I can agree that they set out to test and do show that there is a change in what women CAN detect and rate as relatively more attractive, and that this includes chemicals associated with male sweat or skin, and that this CAN correlate (though we'll see how much) with symmetry. That is of course what they isolated.
What they did not do is attempt to show whether women actually USE these chemicals to discriminate, or COULD USE them in real world settings to select a mate, much less to such an extent that it would have been such a driving factor for genetic selection that it created a relationship for reproductive reasons. Once again, in systems where non-deselection is good enough to maintain existence, correlation does not indicate causation or an organically important relationship, even if one trait also correlates to "good genes".
As it stands I DO have some alternative explanations, specifically in light of their results.
METHODOLOGY
5) Despite claiming that their previous study was small, and that this study would improve upon the earlier study, this group was very small. In addition it was not at all cross cultural, nor cross ethnic. It was also apparently isolated to psych students, who may very well be associated with the authors and their theories. For a study attempting to get at an evolutionary basis for behavior, this is all problematic. Our culture in particular has issues with odors, and sanctioning human odors. This could skew results for different reasons.
6) I have to sympathize with them having to come up with a way of capturing "scent", but I do not have to cut them any slack. Their method was to have people wear a cotton T-shirt to bed for two nights, with a number of restrictions to try to avoid smells coming from outside sources.
They did not provide any explanation of how a T-shirt would be capable of capturing all scents produced by a human body. It specifically cannot pick up scents from breath (saliva), genital-anal area, or feet. Chemicals produced by the body or bacteria native to the body are concentrated in those other areas and produce different chemicals.
In fact T-shirts could potentially favor symmetric individuals as it reaches left-right on the body, rather than across the length of the body.
The fabric might interfere with some pheromones, and some pheromones might simply not last as long as others.
Despite their rules, I notice that they did not mention no masturbation, or even simply touching other parts of the body, followed by the T-shirt. Neither can one discount the possibility of other sexual issues (especially for those with current partners they cannot sleep with) or physical issues (like athletics or current health) which could be in play. And there could also be issues of taking scent "samples" only during sleep rather than during the day when most sweating and pheromone production is likely to go on. It would have been of particular interest to see them explain why that would make no difference.
The plastic bag itself could interfere with certain chemicals, especially when one is measuring scent.
7) I don't really want to get into this particular issue, but its possible that there was an age issue given the larger spread of ages between men and women involved, as well as menopausal issues for women (I don't remember seeing them address these). It might be wrong to place a menopausal or postmenopausal female in the same category as "low-fertility" though it looks like maybe that's what they did?
RESULTS/ANALYSIS
8) Okay this is already getting long, and discussion of statistics might be long and it would definitely be boring. But I need to get across something important.
The pertinent results are listed as r values which in fact are unimpressive, despite discussion suggesting otherwise. For example the high-fert-risk (the ones who should be most correlated) have an r 0.15 for scent attractiveness-symmetry. Thankfully the authors note that that was not significantly different than zero (which means there was no support for their theory at all). After correcting for possible confounding from showers/scents, it only raised to 0.22 which should be of concern to the authors.
In fact the highest they ever get (when they switch to a form of analysis most likely to result in seeing something) is 0.49 which is still not really anything one can say is important. It is there, and it may be internally significant, but it is not "convincing".
They often discuss "significance" but that is actually referring to something about the data to itself and not necessarily toward a conclusion regarding real life correlation.
The best that can be said is that there seems to be a difference in sensing/labeling attractiveness between low and hi fert females.
9) Assuming their own description of r is correct: the results of the first analysis had correlation between fragrances/shower to attractiveness as 0.42, which was extremely higher than symmetry to attr, with a difference significantly greater than that between low and hi fert which they said was meaningful. This raises a question (which they never bring up) of how these supposed pheromones actually have selective effects in the real world, rather than this experimental setting. In the real world other scents will be around and appear to have a much greater effect on attractiveness. This doesn't even have to mean soaps and full baths. How could it come to be selected for?
At this point I have to start wondering if what is actually going on is researchers focusing on something closely and demanding a preference be made, which can be made, but it is in fact artificial. It is not something that would actually take place in the real world.
10) As part of their analysis they say...
That fertile women prefer the scent of male symmetry has now been found in two separate studies (also see Gangestad and Thornhill 1998a). This relationship was also reported by Rikowski and Grammer (1998) from their methodologically similar T-shirt study in Austria involving a relatively small sample of research subjects. When all three studies are considered, the effect of men's symmetry on fertile women's scent attractiveness ratings appears quite robust.
...which makes no real sense. They admit one study has some flaws, they say another is "relatively small" (which if compared to there own is extremely small), and their current one is certainly "small", yet conclude that because results from all three are similar, the ratings are "robust"?
That's not to mention that the additional study likely did not add to cross-cultural issues (Australia is modern and Western). And further this is to take for granted that the results of this study mean much. If the other two studies had correlations as weak as this, we have a "robust" result of "not very statistically important".
11) They do go on to discuss better methodologies and some possible explanations of what chemicals may be involved or how they might be generated. I'll point out what they say would move toward definitive...
Definitive evidence for a scent associated with men's symmetry would involve
the isolation of the chemical and the use of appropriate control odors for eliminating
potential confounds such as novelty. That evidence has not yet been found; hence,
we cannot say with absolute certainty that there exists a chemical emitted by men
that women use to discriminate men who have and have not experienced develop-
mental imprecision.
Yet turn right around and say...
Nonetheless, the current evidence strongly suggests that there is such a chemical(s).
...as if this means something. Well I suppose it might, IF one begins with their assumption as being the only explanation, and IF one believes in the popEP deductive principle that since evolutionary pressures do shape genetics, all things genetic which show a correlation to a potential survival advantage most likely exist due to that correlation.
Without that what can be said at best (giving them full credit for correlations) is that women at high fert appear to be relatively more attracted to chemicals symmetrical men produce, or more repulsed by chemicals asymmetrical men produce (?), when placed in a situation where they must choose based almost purely on that criteria.
Ugh I'm getting sick of this already. It's not like this is my job, and I could be doing more fun things. I think this says enough for what we need here right now.
This message has been edited by holmes, 02-27-2006 09:58 PM
holmes
"What you need is sustained outrage...there's far too much unthinking respect given to authority." (M.Ivins)
™ Version 4.2