Between 1975 and 2010, the prevalence of the word tool or tools in PubMed increased six-fold, to more than 3% of all PubMed articles. If you’re writing about a spade, you’re writing about a tool; if you call a questionnaire a tool, it’s metaphorical. And it’s probably an unnecessary way of writing way, means, or method. A questionnaire, scoring system, or computer program is not a tool. A tool is a solid object (so a computer could be described as a tool, although the essence of a tool is that it is usually simple). You hold a tool in your hands and do things with it. Thus “a simple, reasonably sensitive and specific instrument tool to measure sleep disturbance” is just a way of measuring sleep disturbance. I’ve not invented the phrase instrument tool. I don’t know how an instrument tool differs from a plain tool, but there are quite a few out there.

Tools are stored in a toolbox, and there are a goodly number of them as well. But the toolboxes could mainly be left in B&Q—which, for those outside the UK, is a large do-it-yourself superstore. The study will be easier if researchers “adopt a standard toolbox of measures”? They could more simply just adopt standard measures. Rather than “taking advantage of a program’s parallel processing toolbox to test a variety of parameters”, you could just take advantage of the program’s parallel processing. It’s easy to get confused with metaphor. Snake venoms contain a variety of compounds better avoided, but the compounds have been invaluable to neurobiologists. A review titled “Snake venoms: toolbox of the neurobiologist” was mis-titled. It is the venoms that are the tools.

Are there any simple guidelines to what makes a scientific metaphor appropriate? Apparently, some psychologists like to refer to the brain as a tool, which made me uneasy: if the brain is a tool, who is using it? Alfredo Gaete and Carlos Cornejo feel equally uneasy, and back up their unease in the baldly stated, “The brain is not a tool.” The first half of this quite long paper is metaphysics and theoretical psychology, but they then ask the specific question. They rightly say that there are no absolute criteria for metaphor, because it depends on context. “Language is an eggplant” probably isn’t useful as a scientific metaphor, but it might serve as a poetic one. A scientific metaphor needs to promote further understanding, and favour advancement of theory: Gaeto and Cornejo call this the productivity criterion. The second necessary criterion is that the metaphor is grounded, that is, it isn’t contrary to an axiom of whatever subject it be applied to: Gaeto and Cornejo give the example that a Darwinist cannot suggest explanations that conflict with evolution. They term this the consistency criterion. Both of these criteria lie aside from aesthetics: how “beautiful or elegant or resplendent” the metaphor is.

Gaeto and Cornejo go on to argue that “the brain as a tool” fails both the productivity and the consistency criteria, and I agree. They then rather slyly add that “the brain can be thought of as a tool” [their italics], but that just because you can think of something as something else doesn’t mean that you are right to do so. They make the final sensible statement that inappropriate metaphor risks “loading our theories with expressions that are detrimental to them”.

There are some brilliant scientific communicators, who are well versed in resplendent metaphor. For them, and for me, tool is altogether too prosaic. It passes the consistency criterion, but fails the productivity criterion miserably.

Competing interests: I have read and understood BMJ policy on declaration of interests and declare that my only competing interest is my co-authorship of a book about medical English.

Neville Goodman is a retired consultant anaesthetist and a writer, and co-author of a book on medical English.

Correction: The name Alfredo Gaete was corrected as it was originally written as Alfredo Gaeto.