First, while race is a contentious issue, it is fair to say that it is much more than a social construction. If it were a social construction, then it would be something that we can choose not to have exist. While racial differences are genetically insignificant for the most part (sure, some genes are found more often in certain groups than others, but that is explained in large part through geographic isolation and other historical consequences). On the other hand, race is a legal matter in many cases in the United States, making it much more than an illusory social construction. One drop of "black blood" (sic) has made a person black for at least decades. Even a fairly recent case in Louisiana of a woman who thought of herself as white, had white friends (maybe even a husband), had parents who were thought of as white was listed as black on her birth certificate (I believe that Louisiana is the only state to state the race of a newborn; I may be mistaken), and sued to be recognized as white instead of what was on her birth certificate. She lost (I know about the case through Walter Benn Michaels's book, Our America, you may want to look there for more details).
A social construction also assigns a way of acting to members of a race. In the case of this woman, she "acted" or "lived" as a "white" person but was still ruled to be black. Is it necessary for all members of a "race" to act a certain way, listen to certain music, etc.? Of course not -- our multicultural education has taught us at least that much.
And on a genetic level (from what I've been able to gather; I'm not a scientist) differences between the races are overall very insignificant -- no more than two unrelated people (of a given ethnic or racial group).
How then to approach the dominance of African athletes at short sprints and long marathons?
1) Sprints have been recently dominated by athletes of West and maybe Central African descent, especially by their descendants in other countries (US, Jamaica, Canada, and Great Britain, but other Caribbean islands as well).
2) Long and middle distances (800m and up) by north Africans (primarily the middle distances -- 800m and 1500m/mile -- and east Africans (primarily long distances, 3000m steeplechase, 5000m, 10000m, and marathon), especially Kenyans and Ethiopians.
Jim Fiore states
Distance running is all about conditioning and little about technique. In other words, you can get quite far on raw, natural talent.
He's comparing running to sports like swimming and gymnastics, so his point is largely valid.
Or is it?
I run and know that there are three key training runs for success in middle and long distances: interval workouts (sprints of 200m to mile or even three-mile repeats for marathoners with breaks), tempo runs (run slightly slower than the anaerobic threshold, generally a bit slower than 10k race pace for most runners), and long runs. Each run has a different purpose: teach body to use lactic acid (or to resist it, as most people think), increase body's ability to use oxygen, improve stride/mechanics (there is plenty of "technique" here), hold a steady pace, improve anaerobic capacity by challenging body near race pace, etc., and overall muscle-skeletal development through the long runs.
Short on technique? Perhaps compared to other sports, but in order to train one needs 1) good shoes (there are many kinds of feet; the wrong shoes will cause injuries), 2) a stopwatch (a heart rate monitor would be ideal, but may not be necessary), 3) ideally a track or some fairly flat area whose distances can be easily measured, 4) a place to run. Not a lot, but sometimes not all of these are present. A fairly flat place to run for tempo runs, ideally with distances vaguely known, is extremely useful too.
Just two observations here:
1) What about the difference between men and women? If it is a racial or ethnic genetic trait that is making athletes from West Africa better in the sprints and athletes in the middle distances from North Africa in the middle distances and East African athletes in the long distances, then why is this phenomenon not nearly as notable on the women's side?
Case in point: only two different African women have ever held the record in the marathon (one broke her own record); the record is currently held by Paula Radcliffe, a light skinned woman from the UK (after first breaking the record, she again broke her own record). This year's marathon at the Olympics was won by Constantina Tomescu from Romania -- another European woman (East European at that).
Another example: the 2004 Olympic 100m (women's) was won by a runner from Bulgaria. She's light skinned too.
Asian women have also done well at long and middle distances: the women's record in the 10,000m is held by Junxia Wang of China (set in 1993; I'm assuming there's a good chance she was clean; Jim notes that the record may be tainted; most of the women's records set in the 1980s most likely are, given where runners have been recently it very well may be). The African record is almost 23 seconds slower (but if the record is tainted, then this point is moot).
2) How much of the difference is psychological?
A lot of races, especially the longer ones (but apparently the sprints as well -- look at Asafa Powell's performance at the World Championship last year and the Olympics this year), come down to strategy. Some examples: Wilson Kipketer, a Kenyan born runner from Denmark holds the record in the 800m (the previous hold was from Britain); he doesn't have an Olympic gold despite being the top of his generation. Another example: Alan Webb ran the fastest 1500m and mile time (as well as the second fastest 800m time) last year; finished 8th at the World Championships last year in Helsinki, didn't even qualify for the US Olympic team this year. And I already mentioned Asafa Powell, who, as current world record holder last year at the World Championships, lost to Tyson Gay (who recently ran the fastest 100m time ever under any circumstances -- it was wind-aided).
Why did these people lose their races? In the cases of Kipketer and Webb, it was poor race strategy/tactics. Kipketer was boxed in (surrounded by other runners) and couldn't get around, much less past, them to win. Webb wins a lot of races because he's faster than the other runners. But speed (in 800m or 1500m) doesn't always win -- especially in championship races where the pace is slower (kind of like the last couple minutes of a basketball game), plus by the time of the championship race, they've already run a couple races over the previous days -- few records are set in those circumstances (the last time the 800m record was set at the Olympics was 1976 when a Cuban runner won both the 400m and 800m.
Why did Jeremy Wariner (a white kid, by the way) lose to LaShawn Merritt in the 400m? Poor strategy at the US Trials, after the loss at the Olympics he said his legs gave out -- he still finished second. It could have been poor training (too hard too close to the Olympics), a bad day, psychological (he had lost very few races the past few years), or even a poor lane assignment (he was in lane 7 -- the best lanes are 3 to 5, from what I've heard).
Considering the level of competition, if someone admits, even on a subconscious level, that the runner next to them is more naturally endowed, most of the time, he or she lost the race before the gun sounded.