Scientific Instruments

Aristotelian view

Things behave as they do through the interaction of "natural tendencies." For example, things tend to move toward the center of the universe and that is why they fall. But if there is something in the way, they can't fall. That means that things will not behave as they normally do in forced or artificial situations. That makes experimental manipulation useless as a means of discovery. Experimental manipulation only looks as if it might be useful if you think that things behave, for example, according to the same laws all the time.

Aristotelians did do a lot of observing of how things ordinarily behave. Either way, there has to be some method ("inductive") of going from particular cases to something more general. The dominant version even in the 17th Century was "Aristotle's" principle of simplicity that similar effects, wherever produced are due to the same causes.

Much experimentation in the 17th Century even up to 1900 was of a sort suggested by Aristotelian worries about artificial situations: one big deal in experimenting with electricity was the attempt to produce "artificial lightning." Wilson's cloud chamber, which we now think of as a device for detecting elementary particles, was constructed as a way of studying clouds: it manufactured "artificial clouds."

The distinction between that and experiments directed at laws or idealized phenomena is particularly clear in the case of Cavendish's model electric fish. (Background: electricity, animal electricity, contact electricity.) Cavendish builds a fish-shaped device that is charged up using static electricity. When you touch it, it feels like touching, on the one hand, a Leyden jar (a capacitor), and on the other hand, an electric fish. Today, following Newton, we tend to do experiments to work out laws and then use mathematical models to determine what the results will be in particular cases. Cavendish "built the particular case."

In building experimental systems, certain components become standard. There comes to be, for item x, suppliers of item x. The fact that there are suppliers is important in many ways: you can't supply something that doesn't work reproducibly. Standard devices produce standard calibrations. Standard devices get used a lot.

Reproducibility can be important in getting people to believe in the importance of a phenomenon. (So can size.) There is also an important kind of experiment we haven't even mentioned yet: the lecture demonstration.

Experimenters of the 17th Century

active (air pump, cloud chamber, electrostatic generator), passive (clocks, balances, galvanometer, electrometer, sextant, transit, telescope, microscope), models (astrolabe, armillary sphere, orrery)

-- ShaughanLavine - 28 Mar 2006