Darrel Ince's review of the journal Object Oriented Systems (Multimedia, THES, December 8) headlined "Few flat-earthers on a new horizon", pulled few punches.

Many computer specialists smile when Ince, a prolific author of books on topics as diverse as software maintenance, software engineering, quality assurance, software metrics, practical formal methods, project management, prototyping and the software life cycle, rails against "computing academics whose careers have been built on swinging Tarzan-like from trendy sub-discipline to trendy sub-discipline".

They wonder why Ince, the author of a textbook on formal system specification, publicly castigates a leading journal in the field, and grow sad when Ince, editor of a volume of Alan Turing's collected works, portrays the science inspired, in part, by Turing's vision, as "a graveyard of mathematical posturings".

Non-specialists who relish a good fight, at least as spectators, may smile too. However, even if the nuances of the technical debate pass them by, all scholars can, sadly, recognise Ince's tone: the current insecurity of the academy means that, like humanists who move outside the canon, geographers and historians who think about more than maps and chaps, or thinkers who suggest that, goodness me, knowledge might be a gendered construct, computer scientists who look beyond the perceived immediate needs of industrial practice are a convenient target for lively abuse.

ÁñÁ«ÊÓƵ

Renaissance iconography portrays theory as looking hopefully to the stars and practice staring sadly down at the earth, the former bound and the latter blind. In reality, things are more complicated. Bernoulli, who in the 18th century laid the foundations of the probability theory, which is used in ensuring that complicated telephone networks can handle millions of calls at once, or Boole, who in the 19th century devised the symbolic logic we use in circuit design, devised their abstract mathematical notions to address the problems of gamblers and engineers.

Many of their contemporaries saw in their work what Ince characterises as "no purposive dimension and little validation". Modern banking relies on complicated number theory to ensure the security of automated transactions, just as medical scanners rely on advanced geometry to ensure they correctly reconstruct 3D images of our insides from a sequence of 2D images.

ÁñÁ«ÊÓƵ

Current readings remind us that our notion of science is defined by what we choose to forget as well as what we choose to remember, and not every scientist in the past was a Bernoulli or a Boole, just as not every one of the "European computing academics" who Ince despises will be remembered as a Turing, or even an Ince.

The extraordinary impact of modern computing has generated all kinds of fascinating science, which tries to provide plausible explanations and models of computational phenomena, may shed light on real-world problems and may also sometimes have no conceivable pay-off in our lifetimes. It generates serious debate, about what is worthwhile and what is not, what is and what could be, informed by the people who make and sell hardware and software as well as by policy makers and scholars. It needs lively minds and smart people: it deserves more than tired tabloid invective.

Ursula Martin

Professor of computational science

University of St Andrews