The Duke University Graduate Bulletin describes the Ph.D. dissertation* as a mature and competent piece of writing, embodying the results of significant and original research. Other material available from the Graduate School and items from Chemistry Department files indicate that the style of a dissertation should be that of a recognized chemical journal such as the Journal of the American Chemical Society.
It is conspicuous that these sources avoid any prescription as to content, length, and broad organization of a dissertation. It can be argued with some force that such a statement would be superfluous. A well-honored understanding about the broad outline and general content of a dissertation already exists because the educational aims of writing it are widely accepted. Indeed, the consensus on this point is a very old one and can be assumed to be as international in scope as the Western scholarly community. Thus, the GraduateSchool finds the phrase “a mature and competent piece of writing” sufficient to recall the common understanding.
Was the tradition not viable and enduring, a detailed description of a dissertation in chemistry and other fields would probably be desirable to guide the student.
Considering the situation in chemistry, a comparison with the quite different situation that prevails with regard to the graduate chemical curriculum is instructive. The curricula and the accompanying requirements for a Ph. D. reflect the changing nature of chemistry; differences in educational preparation of entering graduate students, and even changes in educational philosophy.
By implication, the dissertation is a constant factor in the Ph.D. program. Should it be? Can the function presently served by the writing of a dissertation be better performed in other ways? It will be argued that in only exceptional cases will other paths be equally fruitful.
Of the educational goals to be realized by the writing of a dissertation two stand out:
The rapid pace of scientific communication today encourages preparation of journal papers at as early a stage of research as seems justified. How should such documents relate to the dissertation requirement? If the candidate prepares such papers,the first goal listed above will be met. To meet the second goal it is proposed that accepted manuscripts or reprints may properly be considered a part of a dissertation.
The Graduate School requirements concerning style must, of course be followed. It may often still be desirable to add an introductory section and to expand the discussion tersely presented in the paper(s). It would be essential to report all “unpublished” data and to discuss experimental procedures and designs in sufficient detail that a clear account is left for others who will follow.
Finally, it should be noted that the route of paper writing is undoubtedly reserved to the select few.Research directors write most research papers, a process that in no way meets the goals set forth for the dissertation. It also seems likely that one or a set of student written theoretically oriented papers might well be a dissertation. In this case addenda might well be superfluous.