September 16, 2003

Writing To

Digital inscription is a form of displacement. Its fundamental character is to remove electronic objects from the channels of direct human observation. This is reflected in the everyday language we use to talk about the inscription process. The commonplace is to speak of writing a file to a disk; to say writing “on” a disk sounds vaguely wrong, the speech of someone who has not yet assimilated the vocabulary or concepts of computing. We write on paper, but we write to a magnetic disk (or tape). Part of what the preposition contributes here is a sense of interiority; because we cannot see anything on its surface, the disk is linguistically refigured as a volumetric receptacle, a black box with a closed lid. If we were writing on the disk we would be able to witness the residue—even microprint is visible under rudimentary magnification. Instead, the preposition of choice, “to,” becomes a marker for our intuition that the verb “write” is not altogether appropriate, a rough fit at best. The preposition is also a legacy of the von Neumann model, where storage is a physically and logically distinct portion of the computer (the EDVAC, the first stored program computer built according to von Neumann’s principles, used tubes of mercury to store an electrical signal). Writing data “to” the storage element thus entails a spatial as well as a sensory displacement.

We do, of course, often speak of putting a file on a disk. Likewise, “saving to” and “saving on” a disk appear to be used with about equal frequency. Since it is clear that we can thereby conceive of disks or other storage media as a form of material support for data, it becomes all the more conspicuous that we only seldom speak of writing a file on a disk. I would argue that writing implies a level visual feedback that is generally absent from electronic storage media, obtaining instead at the level of the screen or other output device—architecturally distinct components of the von Neumann model. The OED helps chart this lexical unease: in the 1940s, one could comfortably say either write “on” or write “to” tape or disk (or indeed, more commonly, write “into”). Since the 1950s, however, the preferred locution has been simply “to.”

Posted by mgk at September 16, 2003 01:58 PM
Comments

Even more interesting is the phrase, "write to an API", meaning that the code conforms to or uses a specified application programming interface.

Posted by: Mark Bernstein at September 17, 2003 02:24 PM | Link to Comment

Grrr... I had a long comment here, but the window went blank just before I hit post.

How common is "write to" as opposed to "save to/on"? In the handful of computer languages I know, "write" is used in general to mean "create output" (regardless of whether that output is destined to be saved, consumed, displayed on a CRT, etc.). In the 40s and 50s, most people talking about computers would be professional computer experts (programmers or at least operators), so one would expect their language to be influenced by programming syntax. But end users today see "File|Save" in their menus, so "save" is probably a lot more commonplace in the general public.

I suppose another reason to think of digital media as volumetric is because a disk or tape actually does have an outer surface that could be written upon with a marker. Office workers might need to differentiate between the different types of writing.

Posted by: Dennis G. Jerz at September 18, 2003 01:33 AM | Link to Comment

FYI I Googled a bit for such things as "write to" and "write on". The URLs would be unweildy here... if you're interested see http://jerz.setonhill.edu/weblog/permalink.jsp?id=1700

Of course, this doesn't help examine the language of the 40s and 50s. I feel like the guy in the joke who lost his keys in the alley, but he's looking along the boulevard because the light's so much brighter there.

Posted by: Dennis G. Jerz at September 18, 2003 02:03 AM | Link to Comment

Thanks Dennis and Mark, this is useful feedback. Dennis, the only one of your Googled results that surprises me is this one:

* "write on" paper: 273,000
* "write to" paper: 1,300,000

Wonder what accounts for that? (I'd have expected the results there to be flipped.)

Posted by: MGK at September 18, 2003 09:08 AM | Link to Comment

My guess is that the discrepancy between "write to" and "write on" nas nothing to do with the question at hand: Google reports 5,820,000 links for "write to", and only 728,000 for "write on." So, it makes sense that when you throw "paper" into the mix, "write to" would still be more common (though as much more common as when "paper" isn't an added search term). Makes sense - one "writes on" a sheet of paper, but one also uses a sheet of paper to "write to" someone else: a friend, a congressperson, etc. They're kind of apples and oranges, in a sense...

Posted by: Josh at September 18, 2003 11:08 AM | Link to Comment
Due to the proliferation of comment spam, I've had to close comments on this entry. If you would like to leave comment, please send email to me at mgk =at= umd =dot= edu. Thank you.