When, in the first book of A Treatise of Human Nature, Hume wants to exemplify the kind of belief allowing one to make sense of the most everyday experience, he offers an example that has become well known: hearing the door creak downstairs and seeing a porter arrive in his room a few minutes later, he must assume that the man has arrived by way of the stairs. But this example carries on, as the porter delivers a letter

The example of the letter in Hume's hands as an object indecipherable by rational means is one of several moments when the Treatise calls attention to the written word in its tactile form, physically encountered by readers even at the moments when they are asked to suspect material forms of evidence. In order to read a letter, Hume suggests, one must actively make sense of something that is not actually perceptible: forgetting the distance it has travelled (for how could one verify that the posts and ferries still ran, or observe every step of its way?) and the contingencies that inevitably haunt its claims (so much time has passed since they were made), one must handle the letter irrationally in order to make sense of its words on paper. If a text is to be functional and pleasurable, the reader must accept the fiction of the distant friend's presence in writing.

This illustration calls up by inference the task of reading any book, produced, as Hume was well aware, as the result of frequent shuffling across oceans, between publishers and editors, printing, correcting and reprinting.

The eighteenth century was an ‘age of machines’. This much we know. But recently literary critics and cultural historians have become concerned with just how controversial drawing instruments, hot air balloons, mechanical dolls and other products of virtuosi science were in the period. Some thinkers used the analogy of machines and automata to prove that a well-governed, self-propelling society was not only desirable but possible. Others complained that the time and effort that went into exhibiting and selling mechanical trinkets and spectacular exhibitions proved that modern culture was economically corrupt and morally indiscriminate. Nevertheless, already by the first decades of the nineteenth century, machines had become part of the social and economic background. Even those whom we might expect to be against them found ways to accommodate machines and find a moral high ground from which to view them. A Tory naturalist, William Wordsworth could in 1833 look at a railway viaduct and declare:

The claim that the materialist revolution might be assuaged by the sublime is a stretch, but it proves the larger point.

How did this happen? The simple answer is that people just got used to new machines much like we have cell phones and iPods. A better answer has to do with how technologies appear in other technologies, a process that Jay David Bolter and Richard Grusin call ‘remediation’. Normally, the word remediation means to remedy or clean up (as in an environmental spill), but as Bolter and Grusin use it, it also means to resituate one medium into another. Googlebooks is a good example: printed books are photographed and then posted onto the web; the old linear medium becomes subject to the non-linear mechanisms of the computer interface; the limitations of print are thus remediated. The process works the other way as well. No one knew much about cyberspace until William Gibson wrote about it in his novel Neuromancer. Historians of science see something similar at work in the Enlightenment.