Science is founded on principles of objectivity. Scientists seek to be objective in conducting their research, and claim to achieve that objectivity. They often extend the orthodoxy of objectivity to encompass a belief that the account of their work can be objective as well.

Some writers have challenged this complacency. In Darwin's Plots, Gillian Beer of Cambridge University analysed The Origin of Species, finding it to be "a work which included more than the maker of it at the time knew". The context in which Darwin wrote is readily apparent from his description of the "exotic" species that he saw on his voyages in relation to "norms" of English flora and fauna. While not unscientific, these observations inevitably had their root in the larger culture. More controversially, Darwin sought to incorporate elements of pre-existing mythology in his evolutionary theory. He retained the idea of "Mother Nature" for example, while the "tree of life" persists in his branching diagram of the evolution of species.

But The Origin is 135 years old, a book-length work rather than a scientific paper, and written in a florid prose style. It is easy today to read it as a literary enterprise as much as a scientific one. It is correspondingly easy for scientists to continue to believe in objectivity in accounts of contemporary science. "Whatever the scientists' feelings, or style, while working, these are purged from the final work," as the biologist Lewis Wolpert put it in The Unnatural Nature of Science.

Yet an account of a scientific experiment is no different from the record of an epic journey or an historic battle. It is the first step in a myth-making process. There can be no truly objective account of scientific endeavour, however devoutly the scientists might wish, hope, or pretend that there should be one.

³ÉÈËVRÊÓƵ

The scientific style guides admonish against first-person braggadocio and against literary tropes, against entertainment or diversion of any kind, even against historical truth where that might include an account of long, fruitless periods of exploration, or the admission of wrong turnings, moments of serendipity, or lucky accidents.

Yet even from the straitjacket of conventional "objective" writing that is left, the subjective cannot help but erupt. Writing in 1988 in a journal of general chemistry, the Nobel laureate, Roald Hoffmann, imagined the thoughts of a humanist who reads a scientific paper: "She notes a ritual form . . . There is general use of the third person and a passive voice. She finds few overtly expressed personal motivations, and few accounts of historical development. Here and there in the neutered language she glimpses stated claims of achievement or priority." Hoffmann contended that the real content of a paper is the product of a struggle between what should be said in order to preserve the convention of the medium and what must be said in order to persuade readers of the argument and achievement of the authors. "That struggle endows the most innocent-looking article with a lot of suppressed tension."

³ÉÈËVRÊÓƵ

This is true in varying degrees of all the thousands of papers brought forth by the world's scientists to form the body of information known as "the scientific literature", or, more often within a given field, just "the literature". The terminology invites consideration by the techniques of literary criticism. While acknowledging the attempt by scientists to use language in a neutral manner, literary theorists from Roland Barthes to Terry Eagleton have found no grounds to exempt their writing from critical investigation. For all its denial of literary pretensions, the scientific paper is a text like any other.

It is possible to detect hidden content in more recent, and less obviously "literary" scientific writing. When James Watson and Francis Crick revealed the structure of DNA, in a paper in the journal Nature in 1953, they were clearly cock-a-hoop. They wrote in the active mood rather than the customary passive and adopted a lively tone of ironic understatement: "We wish to suggest a structure for (DNA). This structure has novel features which are of considerable biological interest . . . It has not escaped our attention that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material." In this short passage alone, they employed the literary devices of meiosis (the understatement in "of considerable biological interest") and litotes (the expression of a positive idea as the negative of its contrary in "It has not escaped our attention . . .").

Consider the still more recent discovery in 1985 by American and British scientists of a new form of carbon. This chemical element has been known for millennia in the familiar forms of diamond and graphite. It came as a considerable shock to learn that carbon could also form hollow spherical molecules of 60 atoms far more stable than similar-sized lumps of either of these standard forms. The new third form was christened buckminsterfullerene after the maverick American architect of geodesic domes, Richard Buckminster Fuller. As well as bearing a passing resemblance to these domes, the buckminsterfullerene molecule is like a soccer ball with the bonds between its 60 carbon atoms tracing lines like the seams joining the hexagonal and pentagonal leather facets of the ball. It is the most symmetrical molecule yet made or found.

Two documents might be said to contain "objective" accounts of this discovery. One is the log book in which the chemists -- Harold Kroto from the University of Sussex, and Richard Smalley and Robert Curl of Rice University in Houston, Texas, and their students, James Heath and Sean O'Brien -- recorded their observations. This is the only contemporary document that could form an authentic record of what happened in the laboratory minute by minute and day by day. Kept as it was by a rota of the students, it is, however, incomplete and ambiguous. It is in its turn merely a commentary on the computer data files that held the experimental results in the graphical form of mass spectra.

The only other record that has some claim to objectivity is the letter announcing the discovery in the journal Nature. This text is important not just because it was the first announcement to the world of this new form of carbon, buckminsterfullerene, but because it will remain for all time the primary published source available for independent study. If objective, then, unlike the log book, it is very consciously so.

The paper is short, well written by the standards of scientific papers, and contains little jargon. It is the only description by the people who made it of one of the most important discoveries in chemistry this century. To read it is to share the excitement of the act of discovery in a way that cannot be recreated by an outsider.

³ÉÈËVRÊÓƵ

Its title provides the first clue that something unusual is afoot: "C60: Buckminsterfullerene". This is a clever formulation. Despite its length, the name demonstrates a certain economy. It intrigues the casual reader --and, importantly in a general science journal such as Nature, the non-chemist and even the non-scientist.

Next comes the abstract of the paper, the first sentence of which makes up in completeness what it lacks in elegance: "During experiments aimed at understanding the mechanisms by which long-chain carbon molecules are formed in interstellar space and circumstellar shells, graphite has been vaporised by laser irradiation, producing a remarkably stable cluster consisting of 60 carbon atoms." This brutally concise summary of two weeks of experimentation is significant for its suggestion of serendipity at the very outset of the paper: we were doing A, when we noticed B . . .

³ÉÈËVRÊÓƵ

The authors go on to discuss what kind of 60-carbon atom structure might be so stable: "We suggest a truncated icosahedron, a (polyhedron) with 60 vertices and 32 faces, 12 of which are pentagonal and 20 hexagonal." The use of the active -- "we suggest" -- is comparatively rare in scientific literature and is frowned upon by some stylists. It provides some indication of the excitement of the discovery that is to be described and of the pride felt by the discoverers.

Nature has strict limits on the number of figures that may be shown with a paper, but these authors give their first figure over to, of all things, a photograph of a soccer ball, which has the same form as the molecule they are describing. It is a populist gesture where a conventional diagram would have been more usual. As they get into the meat of their paper, the scientists' excitement becomes palpable in quick-fire phrases laden with jargon: "Carbon clusters were expanded in a supersonic molecular beam, photoionised using an excimer laser, and detected by time-of-flight mass spectrometry." It is almost like a cricket match commentary: the ball was thrown, struck, and caught. Zip! zap! splat!

Moving on from the description of the experimental method to discussion of their results, the authors aim to carry the reader along with them in trying to explain the 60-atom carbon cluster in terms of the known forms of carbon, diamond and graphite. But both forms fail to explain it. They reinforce their new molecule's claim to specialness by commenting that "this result seems impossible". We have been led through the problem only to find it intractable. Our expectation is building for a surprise denouement: "Only a spheroidal structure appears likely to satisfy this criterion . . . An unusually beautiful (and probably unique) choice is the truncated icosahedron depicted in Fig. 1 (the soccer ball)." The fact that they made a "beautiful choice" is tacit admission that aesthetic intuition guided that important last leap of the imagination toward the final structure.

At the conclusion of their paper, having divulged their news, the authors adopt a lighter tone: "We are disturbed at the number of letters and syllables in the rather fanciful but highly appropriate name we have chosen in the title to refer to this C60 species. . . A number of alternatives come to mind (for example, ballene, spherene, soccerene, carbosoccer), but we prefer to let this issue of nomenclature be settled by consensus."

But they protest too much. They are not so "disturbed" that they do not go ahead and name it buckminsterfullerene anyway. The fact that they used this new name for the title of their paper is ample indication that they wish it to prevail.

The idiosyncrasy of their choice has since been vindicated. The discovery of buckminsterfullerene or "bucky balls" has done much to rejuvenate the field of chemistry which had been perceived as less glamorous in recent decades than particle physics or molecular biology. The choice of name, meanwhile, has ensured the popularity of the discovery well beyond the world of chemistry.

³ÉÈËVRÊÓƵ

Hugh Aldersey-Williams's book, The Most Beautiful Molecule: An Adventure in Chemistry, was published by Aurum Press on 17 November, price Pounds 18.95.