Editing the Past

I’m looking into ways of making my posts living, breathing things. I don’t feel like I have any of the answers yet, so I’m compiling some questions about the effects of editing the past.

Looking at the archive…

As an evolution of feelings

What’s the best way to update old blog posts? Does my audience care if I’ve changed the meaning of something from 2011 or simply removed the harshness of my tone? What about the fact that I often write to get it halfway there (or even 5% there), just to help myself understand my feelings?

What if my feelings turn inside out? What if I want to delete something, because it’s just plain bad?

Do I get to decide what’s important as the author? Why? Or do my readers get to decide with traffic, comments, or attention?

Is frequency important, or is all of this an exhaustive attempt at making order out of chaos? Maybe I should stop right now.

As a written record

What are the ethics of deleting something? Or hiding it? What if I just shove it in a corner or an armpit, only to be found by Google or someone with a link? Why bother keeping it there, if it’s not worth sharing openly?

Is it an archive if I don’t preserve my words as they were originally posted? Or does it break the web to think anything should be static for more than a month or two? Do we breathe here in minutes, months, or milliseconds?

As a resource

Is this helpful if it’s not updated? Should I announce every change I make, or put notes within each article? Do I get to summarize my own summaries, or is there a more programmatic way of showing changes, like a differential or commit?

Should I notify anyone of anything? Or is it annoying and uninteresting to know about teeny changes on a personal site? But isn’t everything connected? Even if my opinions are small and unsharpened, isn’t that the point of sharing them and working on them over a lifetime?

So many things are making me think about this, not least of which is the Contents Book Club. Join us in reading The Library at Night.


  1. I like this kind of long-term thinking. Personally, I’m taking a two-part approach.

    I set up an “errata” page on my site, where I list changes, a practice borrowed from old book printers. Posts to this section do not show up anywhere else in the site (nor even on the original post) but they’re there if people care to look for them. (And if I make a spelling/punctuation/grammar fix I don’t bother noting it anywhere.)

    In cases where I have radically changed my opinion on something or where more substantive alterations are in order (there are many when you’ve had a website for 15 years), I plan to create a mechanism for adding a large red asterisk to the blog post’s title, and add an emendation at the bottom, or a link to a second post explaining the change. I haven’t gotten around to implementing this part yet.

  2. I’ve been tackling this problem for the IAI as well as my own work. What I have planned is to tag all resources with an extent, which represents an envelope in space and time, and use that to inform not only what gets seen and where (though I’ll have to see a good chunk of the corpus tagged that way before I can say how that information will be used), but also what needs to be updated and when (which I’m more clear about).

    Putting a temporal extent on content essentially gives it an expiry date, or explicitly doesn’t. Both ideas are useful. If I write about a current event, it will probably no longer be interesting in a month or even a week, outside of a purely historical context. (It’s one of my main criticisms of blogs—that they’re overly concerned with the now, and organizing them by publication date invariably masks older content which might be more relevant than that which is newer. It’s also why I chose to leave the dates off of the essays on my site, so readers can’t tell how old they are. The danger there, of course, is that some have invariably lapsed in relevance.)

    The next necessary tool is a synoptic view: something to abridge the task of keeping content current. Something that shows attributes of all the content all at once, and enables the filtering of resources based on those attributes, including extent. Those which have expired, or are about to expire, in particular. The resources which lack an extent for some reason, or have been explicitly marked “eternal”, are interesting subjects in their own right. I don’t know of anything that does this, at least not to my specifications. So I’m making one.

    Finally, there is memory. Specifically, version control. Unfortunately at the moment all version control systems (except for maybe darcs) seem to operate at the level of the file or conceptual equivalent. But a file is really a relation between a name (and location) and an arbitrarily-long, anonymous sequence of octets. But there are other types of relation, and the octet (read: character) sequence can also be as short as one.

    The idea is significant because of our bad habit of treating documents and files (or other named-blob constructs) as equivalent. When you alter a file, even changing a single word, you essentially mint an entirely new document, destroying the old one in the process. It’s like erasing pencil marks without so much as leaving the slightest palimpsest of what was written before.

    Our changes are destructive by default. We need not only extra effort and machinery, but courage to keep them around. Your questions are questions I’ve been thinking about for a very long time, and they have governed my decisions in every aspect of my own work.

    I really appreciate that you’re thinking about this. It gets my own mind going.

  3. This is interesting… How do you decide something should be “retired” as you say? What if someone was referencing that link in an article or presentation and they want to give you credit? And what if you want to reference it later, for your own purposes? Do you trash the file or make it private?

  4. I only “retire” a post if I deem it to have almost no value, informational or otherwise, a duplicate of something posted elsewhere, poorly written, etc. I’m 99% certain no one has ever linked to one of these retired posts. Due to their low quality they never get that kind of attention. Since I use a CMS I just mark the post as hidden, so it’s still “there” in case I ever need it again. I don’t like doing it, and obviously the goal is never to put anything up that I might one day retire. I think there were isolated incidents over the last fifteen years where I was just lazy, and “retiring” allows me to clean those out in a transparent way.

Comments are closed.