NEWS AND NOTES LETTERS TO THE EDITOR
Dear Sir: Your editorial was called to my attention by Dr. J. A. Moore of the University of Delaware chemistry department. Two comments particularly stand out: “Unfortunately, most of the philosophers ignored a very basic t e m t of philosophy: to define terms. From my mass of notes, I have been unable to extract a meaningful definition for ‘philosophy of technology.’ It was quite obvious almost from the very beginning that the philosophers and scientists marched to different drummers.” And: “I am all for developing a philosophy of technology. But I am not sure I am willing to let the philosophers do it for me. Certainly not the political and moral philosophers.” As the arranger of the conference, my first temptation was to react angrily. You and your colleagues were-very kindly, I think-invited to sit in on what was originally planned as a conference of philosophers for philosophers. I thought it was clear enough from the brochure-invitations that ours was an attempt to open up, within philosophy, a new area of investigation that would be academically respectable. I was thus tempted to reply: What right does he have to tell philosophers what they can do-even with respect to technology or its practitioners-any more than philosophers have a right to dictate what members of the scientific community should do? Again there was a temptation to ask whether, for scientists, to moralize on a subject means to dictate rather than to persuade. On second thought, however, my reaction is much more temperate. In one sense you are right: philosophy of technology ought not be left to philosophers alone. This should be an interdisciplinary field, possibly even along the lines suggested: “Possibly those we should seek for this study are those who appear as philosophers to scientists . . . and as scientists to philosophers.” Perhaps what went wrong at the conference, if anything, was that in our moralizingwhich, incidentally, is not all we did; there were elaborate attempts at definition-we did not succeed in persuading our working-scientist colleagues in the interdisciplinary effort we were attempting to get off the ground. There will be more of these conferences, the next a t the AAAS meeting in Boston in February 1976. Let’s hope that our friends in the technical community will not be turned off by one effort to persuade which they deemed a failure; that they will come back again to join with us in what we see as an important venture for twentieth-century man that transcends the narrow limits of academic philosophy.
Paul T. Durbin Department of Philosophy University of Delaware, Newark, Delaware 19711 Received September 26,1975 Dear Sir: In Vol. 15, No. 2,1975 of this journal, Edward G. Mazurs suggested that a numerical identifier for the elements should be based on the Janet form of the periodic table rather than on my proposed numbering scheme.’ (From this point on, for convenience I shall call his proposal the J-scheme and my proposal the L-scheme.) I agree that the J-scheme has the advantages described. However, the numbers of the L-scheme, although originally based on a classical periodic table, were chosen as they are now expressly to achieve certain conveniences of operation and to lessen certain disadvantages of any scheme based directly on a classical period table. The L-scheme uses a two-field identifier and so requires 33% less storage space than the three-field identifier of the 252
J-scheme. This could result in a significant saving of space and execution time in the processing of large files. Another advantage of the L-scheme is that its numbers, although not physically familiar, were chosen to suit a variety of chemically natural unions and distinctions of element sets. The NGRP and NROW values emphasize the chemical analogies commonly seen as existing between the representative and transition elements without allowing extensions of those relationships to the inner transition elements. For example, in the J-scheme the inert gases NeAr-Kr-Xe-Rn, the transition elements Fe-Ru-Os, and the inner-transition elements Sm-Pu all have the same electron number, e = 6. This is physically realistic but chemically not very useful. On the other hand, in the L-scheme, these sets have NGRP = 1, 9, -10, respectively, allowing immediate distinction. As another example, the J-scheme has e = 5 for groups VIIA and VIIB, allowing the proper analogy between these subgroups to be recognized, but the inner transition elements Pm-Np also have e = 5, a fact of little chemical use. On the other hand, the L-scheme has NGRP = 8 for groups VIIA and VIIB and yet NGRP = -11 for Pm-Np. Similar distinctions and analogies exist for other vertical and horizontal sets and set unions of the elements. As a result of these patterns, the L-scheme often allows a more rapid screening for element sets than the J-scheme. T o show this, L-scheme and J-scheme criteria (in the form of Fortran logical variables) were set up for the recognition of 34 commonly distinguished sets of elements. (An example of such criteria, for the Cu-Ag-Au subgroup, is: L-scheme: NGRP.EQ.2.AND.NROW.LT.10; J-scheme: L.EQ.2.AND.E.EQ.9; thus each criterion requires two field tests.) The L- and Jschemata require the same number of tests for 18 (53%) of the sets (each representative row 1 through 4; groups IA through VIIA, IB through VIIB). The J-scheme requires fewer tests than the L-scheme for 3 (9%) of the sets (representative; transition; inner transition). The L-scheme requires fewer tests than the J-scheme for 13 (38%) of the sets (each representative row 5-7; inert gases; group VIII; each group VI11 vertical subgroup; each transition row 1-3; rare earths; actinides). Moreover, the L-scheme allows 18 (53%) of the sets to be recognized with a single field test (representative; each representative row 1-7; each transition row 1-3; inert gases; group VIII; each group VI11 vertical subgroup; rare earths; actinides), whereas the J-scheme allows only 7 (21%) of the sets to be so recognized (representative; transition; inner-transition; each representative row 1-4). In addition, the L-scheme makes explicit provision for the coding of element types by the use of zero value fields for NGRP or NROW; the J-scheme makes no such provision. In conclusion, I suggest that when storage space and search time are not important factors and where the use of natural parameters is of value, the J-scheme may be the better choice. On the other hand, when large files are being processed and storage space and search times are critical, the L-scheme offers definite advantages in spite of its use of unnatural code numbers. Loach, K. W., “A Numerical Identifier for the Chemical Elements, Expressing Their Periodic Relationships”, J . Chem. Doc., 14,198 (1974).
Journal of Chemical Information and Computer Sciences, Vol. 15, No. 4, 1975
K. W. Loach Chemistry Department SUNY College at Plattsburgh Plattsburgh, New York 12901