[In The Golden Bough Sir James Frazer deduced] that in essence primitive magic was not like primitive religion, as most observers had assumed, but was instead similar to science, in its belief that the universe was subject to "immutable laws, the operation of which can be foreseen and calculated precisely". The Golden Bough makes this claim overtly…. [And] it is a relatively short step from saying that magic is very like science to saying that it is actually a form of science. It is this further step which many science fiction authors have, with varying levels of seriousness, been happy to take. (pp. 121-22)

[The] real potentials of the "Frazerian" story were exposed as well as anywhere … by Fritz Leiber's unduly-neglected novel, Conjure Wife…. (p. 122)

Leiber's hero, Professor Saylor, discovers suddenly and by accident that his wife has constructed round him a great web of magic defences to cover him from the malice of the other faculty wives, all of whom, like her, are witches by instinct and tradition. Dismissing it as superstition, he makes her burn her charms; and then, of course, his life turns into a paranoid's nightmare, with student accusations, missed promotions, charges of academic plagiarism, and so on. In the end his wife, left magically defenceless, is turned into a soulless zombie by her female enemies…. [While Leiber's] images of the powers of witches are at least as gruesome as ancient ones, he nevertheless accepts magic as ethically neutral, usable protectively as well as aggressively. Nor does the place of magic against religion concern him at all, however vital it was for the witch-hunters…. Further, at the moment of crisis The Golden Bough appears, as talisman-cum-guidebook. For Saylor is a professor of sociology (which we would now call social anthropology) and, faced by a zombie wife, he falls back on his academic speciality. He accepts the assumption that the superstitions he has studied detachedly for so long are all garbled reflections of a real truth; takes down his textbooks (The Golden Bough is the only one mentioned); finds some seventeen formulas for calling back the soul recorded by primitive peoples, and reduces them all to a master formula. This combination of superstition and scientific method proves unconquerable, and the story ends … with triumph and reunion.

The surprise in all this, for an unprepared reader, lies in the direction of Professor Saylor's progress. When we hear the word "magic" we inevitably think of reversion, savagery, effortless absence of ratiocination; to find magic then put into an academic context and sharpened by mathematical rigour is inevitably arresting. The juxtaposition becomes part of the stock-in-trade of all "magical" authors, who have a particular penchant for setting stories in and around learned conferences; there is an especially close resemblance to Conjure Wife thirty years later in Roger Zelazny's Jack of Shadows …, where a computer is even dragged in to replace Professor Saylor's symbolic logic. But it is obvious, too, that Leiber enjoys the process of academic argument for its own sake. After all, if his hypothesis is true, it throws up one major question straight away: why has magic never been reduced to order before (given the amount of research dedicated to it the world over)? The question is in a way the reverse of one asked by Frazer, which was why magic had not been exposed before: and the answers to both are curiously similar…. "Magic is a practical science", Saylor theorises, because it is inevitably concerned with "getting or accomplishing something". This means that the personality of the operator is a part of the magical operation; and this means that experiments are inherently non-repeatable. One of the bases of scientific method is accordingly removed, helping on the one hand to explain the absence of any "general theory" of magic, and on the other administering a check to modern assumptions about the universal scope of experimental science. A second point returns one to the definition of magic in the [Oxford English Dictionary], "the power of compelling the intervention of spiritual beings". Obviously, if personalities rather than forces are the object of experimentation, further irregularities are likely to enter, making the whole thing more difficult. And finally Saylor notes that "Magic appears to be a science which markedly depends on its environment"—in other words, it is subject to rapid change. The constants of physics may perhaps change as well (so Leiber suggests), but if they do, they do so slowly. Magic, however, needs to be continuously updated by trial-and-error, and is as a result likely every now and then to fail and be discredited. (pp. 122-23)

Conjure Wife draws power from its cool and rational tone, its everyday setting, while its central images—the cement dragon, the Prince Rupert drop, the shattering mirror—all carry a physical as well as a magical explanation. The book's penultimate paragraph, indeed, offers a rational explanation (that all the women involved are psychotic) as an alternative to the fantastic one (that they are all witches), while the last words of all are Professor Saylor saying evasively "I don't really know". All this makes Conjure Wife fit one rather strict definition of fantasy, that it takes place just as long as one is uncertain about how to explain events. However, it also points out one way in which Conjure Wife does not fit the normal development of 'Frazerian' science fiction, for all its pioneering motifs and explanations.

This is, that most 'worlds where magic works' are alternate worlds, parallel worlds, future worlds, far-past worlds. Conjure Wife is one of very few to be set in a recognisable present. It gains from this, of course, in realism; but loses inevitably a quality of romance. It has witches, and spells, and even the glimpsed presence of He Who Walks Behind; but there are no centaurs, or werewolves, or mermaids, or basilisks, or any of the other ancient images of fantasy. The only dragon in Conjure Wife is a cement one. Yet there is clearly an urge in many writers and readers to resurrect these images and use them again, partly no doubt as a result of 'escapism', but at least as much out of a kind of intellectual thrift: ideas compulsively attractive to mankind for so long, it is felt, are too good to throw away. Nevertheless this urge, powerful though it is, is met by an equally powerful current of scepticism. Twentieth-century readers, especially those with some scientific training or inclination, cannot even pretend to believe in anything that makes no sense, i.e. anything that has no rationalistic theory to cover it. Frazer and The Golden Bough provided a rationale for magic, as exploited by Leiber in Conjure Wife. (pp. 123-24)