13 



Under these conditions the scientist is — or may be thought to be — 

 subject to the temptation of advising not only on what but as to 

 whether. Even when a survey of a field of science concludes that more 

 money should be spent on research, it is in effect making a value 

 judgment on social priorities. Another danger is that in the interpre- 

 tation of data to the poUtician the scientist wall incUne, or permit him- 

 self to be encouraged, toward a finding in response to the situation 

 rather than strictly hmited to the data. Scientific data are rarely abso- 

 lute, and are more usually approximations or probabilities. Informed 

 judgment plays an important part in scientific decisions. Here the 

 politician may be at the mercy of the scientist, but at the same tinae the 

 politician — by structuring the situation — may influence the inter- 

 pretation of the data. 



The difficulty is compounded by the difference in symbols of the 

 scientific and the political subcultures. Communicating among them- 

 selves, scientists express technical issues in confidence levels, figures 

 of merit, uncertainties and probabilities, all in quantitative terms. But 

 in communicating with politicans, they encounter the obstacle that 

 the accustomed quantitative indicators lack significance and need_ to 

 be translated into words. For example, the difference between a purity 

 of aluminum of "four nines" and "six nines" is understandable to a 

 scientist, but not necessarily to a politician. (These are 99.99 percent 

 pure and 99.9999 percent pure.)^ Both forms of aluminum are very 

 pure, but one is more so than the other. The communication could be 

 improved by giving relative costs per pound of material at each level 

 of purity (say, 40 cents and $800). Or by indicating how much of 

 each is normally produced (say, several tons versus a few poimds). 

 Or it might be helpful to indicate the kinds of uses each level of purity 

 has. The purpose of scientific language is to concentrate meaning. 

 The necessity to explain and clarify his terms leads the scientist to 

 virtually endless explanation, and it is sometimes easier to accept 

 shortcuts in which the truth is somehow lost. 



« Actually, for higher orders of purity the scientist would be unlikely to use a percentage figure for total 

 Impurities, except as a loose, order-of-magnitude term, measured by electrical resistivity ratio. If purity 

 ■was important for some scientific test, it might be statisticaUy determmed throughout a sample, for five 

 or six specific impurities, but not for others. Levels of some impurities would be important for the scientist s 

 purposes; others would not. 



