Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!ccu.umanitoba.ca!frist From: frist@ccu.umanitoba.ca Newsgroups: sci.bio Subject: Re: Environment vs. Genetics (was: Evolution vs Genes ) Message-ID: <1991Jun24.182627.595@ccu.umanitoba.ca> Date: 24 Jun 91 18:26:27 GMT References: <3494.28656C72@ofa123.fidonet.org> Organization: University of Manitoba, Winnipeg, Canada Lines: 93 In article <3494.28656C72@ofa123.fidonet.org> Kirk.Reeves@f450.n101.z1.fidonet.org (Kirk Reeves) writes: >Recently on the Donuhue show, I saw this bioliogist claim Orietinal are smarter >than whites, whites are smarter then blacks, but Black have a bigger sex drive. >Being black, I totally disagreed with this man. I was wondering since the >people here are scienists (real scienists) if the question of evolution vs genes >has been quoted: Solved. >The best answer (and the fairest) was a book written in 1980? called the >Genesisactor. What I was wondering is amont bioliogists, has that book and the >theories (It not genes or evolutions it just your Basic better man) >has been disputed. THank you > > >-- >Kirk Reeves >Internet: Kirk.Reeves@f450.n101.z1.fidonet.org >Compuserve: >internet:Kirk.Reeves@f450.n101.z1.fidonet.org >-------------------------------------------------------------------------- This is an old problem that was solved long ago by geneticists. The phenotypic variance Vp (the variance in a trait that we can measure) is equal Vg (the variance due to genetic factors), plus Ve (the variance due to the environment.) The only problem that arises is when yahoos with an axe to grind make sweeping generalizations that are totally unjustified given the data at hand. Although whole books have been written on this subject, let me just make a few relevant comments: 1) First of all, the trait must be straightforward to measure. Good quantitive traits include %fat in milk, height at maturity, yeild in kilograms per hectare, and so forth. Things that are hard to define, such as intelligence or sex drive are bad examples of quantitative traits. Yes, you can assign precise definitions to such traits (eg. IQ scores) and plug these numbers into a computer and get an answer, but you can conclude almost nothing from it, because the definition itself is inadequate to truly describe such fuzzy concepts as intelligence. 2) How you set up the experiment can have profound consequences on the outcome. Let me give an example. Suppose we wished to determine whether or not blacks, as a group, like watermelon more than other groups. How do we do this experiment? Do we examine the buying preferences of different groups, and see who buys more watermelon? Economic or marketing factors could bias our results. How about conducting a survey? How you write up the questions can bias the results. Go to shopping malls and hand out free samples to passersby? This experiment might test the prediliction towards participating in marketing tests, rather than fruit preference. Okay, have a choice of fruit available. But will the selection available, or the quality of each sample make a difference? Social scientists tear their hair out over problems like this. 3) Quantitative traits are distributed within a population around some mean value. Think of a bell-curve that might represent this distribution of scores. The variance of this distribution can be visualized by thinking about how broad the curve is. Statistically, population variance is defined as the sum of the squares of the deviations of each member of the population from the mean, divided by the size of the population. The important thing here is that individuals in populations vary, for any quantitative trait. Furthermore, the bell-curves for two populations are likely to have substantial overlaps, for the type of traits we're talking about. Thus, while you might be able to show that the MEAN scores for a particular IQ test are Oriental > White > Black, that says nothing about distributions, and certainly does not say that ALL orientals are smarter than ALL whites etc. Furthermore, you have the problems of experimental bias and definition of the trait, as mentioned in 1 & 2 above. 4) The components of variance are VERY hard to measure experimentally. Plant breeders will often go through laborious crosses to create inbred populations in which the genetic component of variance is eliminated, in order to measure environmental variance. You don't have this luxury with people. 5) The heritability of a trait is roughly defined and the genetic component of variance expressed as a fraction of the total phenotypic variance. (I'm oversimplifying here, but it serves for our purposes.) Geneticists have found that the heritibility is strongly affected by environment. Thus, in one environment, phenotypic variance might be mostly due to environmental influences, while in another environment, genetics might be more important. So WHERE you do the experiment can greatly influence the outcome! The take-home lesson here is that quantitative genetics is very hard to do, and even harder to do so well that you can get results that are worth anything. Put another way, it's easy to plug some numbers into a computer program. It's hard to design and execute a well thought out experiment. =============================================================================== Brian Fristensky | Department of Plant Science | Freedom begins when you tell Mrs. Grundy University of Manitoba | to go fly a kite. Winnipeg, MB R3T 2N2 CANADA | frist@ccu.umanitoba.ca | Office phone: 204-474-6085 | - Robert A. Heinlein FAX: 204-275-5128 | ===============================================================================