Lind’s reputation as an experimental nutritionist rests mainly on his classic experiment of 1747 in which he compared the potencies of a number of supposed anti-scorbutic remedies (cures for the vitamin-deficiency disease scurvy). Lind’s experiment had obvious commendable features. He selected six groups that were as similar as possible at the beginning of the experiment and maintained them throughout under the same general environmental and dietary conditions. The groups differed from each other only in respect of the type of treatment used. Five groups of ailing seamen failed to respond to their supposed cures; the sixth group, which received a dietary supplement of oranges and lemons, recovered from the disease.
At first sight this would appear to be an early example of the type of scientific procedure proposed by Descartes in 1637 for eliminating all but one of a number of possible relationships. In the ideal type of “critical” experiment, however, each “possibility” should be derived from existing data (or be a logical extrapolation of it) and the whole should be structured so that alternative possible explanations are excluded. Failure to satisfy these requirements reduces an experiment to a level of controlled empiricism. This was the weakness of Lind’s work; the six “possible” cures that he compared were presumably selected empirically from those currently favored by ships’ surgeons; he gives no indication that his choice was governed by any other consideration. His experiment “succeeded” simply because one of the “remedies” contained vitamin C (the anti-scorbutic factor) whereas the other five did not.
It is interesting to speculate what effect a different choice of “remedies” would have had on the course of events. Had Lind used, in place of his oranges and lemons, a sixth remedy devoid of vitamin C-say the mineral waters favored by writers such as Linden, or Bishop Berkeley’s tar-water cure-then all six groups would have given a negative result. On the other hand, Lind could well have used six remedies all of which contained vitamin C. In 1745 John Wesley published his Primitive Physic, a popular manual of remedies, which by 1791 had reached its twenty-third edition. Suppose that Lind had selected six remedies from Wesley’s list of eleven anti-scorbutics. It is virtually certain that all six groups would have recovered, and the experiment would have done little more than confirm the observations of Wesley and others.
At first sight this would appear to be an early example of the type of scientific procedure proposed by Descartes in 1637 for eliminating all but one of a number of possible relationships. In the ideal type of “critical” experiment, however, each “possibility” should be derived from existing data (or be a logical extrapolation of it) and the whole should be structured so that alternative possible explanations are excluded. Failure to satisfy these requirements reduces an experiment to a level of controlled empiricism. This was the weakness of Lind’s work; the six “possible” cures that he compared were presumably selected empirically from those currently favored by ships’ surgeons; he gives no indication that his choice was governed by any other consideration. His experiment “succeeded” simply because one of the “remedies” contained vitamin C (the anti-scorbutic factor) whereas the other five did not.
It is interesting to speculate what effect a different choice of “remedies” would have had on the course of events. Had Lind used, in place of his oranges and lemons, a sixth remedy devoid of vitamin C-say the mineral waters favored by writers such as Linden, or Bishop Berkeley’s tar-water cure-then all six groups would have given a negative result. On the other hand, Lind could well have used six remedies all of which contained vitamin C. In 1745 John Wesley published his Primitive Physic, a popular manual of remedies, which by 1791 had reached its twenty-third edition. Suppose that Lind had selected six remedies from Wesley’s list of eleven anti-scorbutics. It is virtually certain that all six groups would have recovered, and the experiment would have done little more than confirm the observations of Wesley and others.