Something you allude to, but don’t make very explicit, is that legibility occurs inside the reader. In this way, it is much like beauty. What we term beautiful things are things that either we personally find beautiful, or things that most viewers will believe are beautiful (out of what subset of observers?). That said, I expect the tips in “How to be Legible” will tend to improve legibility for nearly all readers, so I definitely think it’s coherent to talk about a work as legible in and of itself.
An aside: I initially interpreted “term of art” as something like “a term which is used for artistic flourish”, since I was unaware that it had a definition as a phrase (I thought it was, under my naive definition, a term of art). I was slightly saddened by the fact that it isn’t an industry term (and therefore is not, according to the dictionary, a term of art).
I expect having a handle with which to say “no I don’t have a concise argument about why this work is wrong, and that’s a fact about the work” to be very useful.
That a work is Epistemically Legible doesn’t mean you’ll comprehend it: you may be lacking necessary background context, for example. See the section Legibility vs Inferential Distance.
In this case, an E-Legible work will still bless you with the awareness that you were missing background context, so that you didn’t understand what was said—as opposed to giving you a fake feeling that you understood. That’s more a property of the work than of the reader, right? You cannot sensibly discount the work’s contribution to such a result.
To be clear, I do believe that the work itself is a significant contributor to the epistemic legibility. My point was merely that the work can only ever be legible with respect to a particular reader or audience (or even expected audience). In this way, I believe it is similar to inferential distance. An idea is not inferentially distant on its own, it is only inferentially distant from someone’s perspective[1]. Where inferential distance deals with the difficulty of understanding a work, epistemic legibility deals with the difficulty of verifying a work. Perhaps an example in which the inferential distance is low but epistemic legibility is low or high depending on the audience will be illuminating:
Rajiv writes a well researched book about encryption algorithms, in which he cites various journal articles published on the subject. Carol is a bright computer scientist with poor judgment who phished people for their bank credentials and got caught. As a result, she is imprisoned and due to the nature of her crime forbidden from accessing the internet while serving her time. She receives a copy of Rajiv’s book through the prison library system, and enjoys it quite a bit. She understands all of the ideas, but wants to check out some of the journal articles to see if Rajiv relayed the ideas correctly and also to further her understanding. The prison librarian checks the price of getting access to these articles, her eyes briefly widen when Carol informs her that she intends to get about a dozen of these articles, and then the librarian tells her there is no way in which she gets any of these articles. Carol is completely unable to verify any of the claims made in Rajiv’s book. A year later, Carol is released from prison, and a couple quick searches reveal the journal articles that Rajiv references, and she is able to verify Rajiv’s claims to her heart’s content. In this example, the inferential distance is low throughout, but the epistemic legibility of Rajiv’s book depends in part on whether or not Carol is imprisoned (unimprisoned Carol could be replaced with some other bright unimprisoned computer scientist).
That was an example of there being differences in the access of information that varies the epistemic legibility of a work. Epistemic legibility might also vary depending on audience when there is the potential for a double illusion of transparency. When this occurs, some audience members will walk away with a confident but incorrect understanding of the work while others might walk away with a confident but correct understanding. This will frequently occur when the language used has variable interpretations (which Elizabeth covered with her discussion of “common knowledge” and inferential distance) or when inferential distance is high.
I don’t think this point is central to Elizabeth’s post (which evaluates works from the perspective of a static reader), so I apologize if I had given the impression that she made some error by not placing emphasis on this. I merely thought it was an interesting frame from which to examine the idea of epistemic legibility.
One point of contrast is that inferential distance likely varies more between audience members than epistemic legibility for a given work. This difference in variability means that it is generally more useful to talk about the epistemic legibility of a work as a whole, but to focus on individual differences for inferential distance
Thanks for the detailed example of what you meant, it’s really helpful. I see what you’re getting at, and it would be insane to say legibility was completely independent of the reader, but I do view epistemic legibility as much more dependent on the work and not the reader than inferential distance, where the load is basically balanced.
In my model, Carol’s ability to check out the articles does not reflect on the book’s epistemic legibility. It definitely affects her ability to check or understand the work, but I’d say the book did its work in being legible by making it possible for Carol to figure out how she needed to follow up, or even just that follow-up was necessary. She’ll understand more when she can check out the articles, but the book met its bar of legibility by not giving her a false impression of understanding more than she did.
Something you allude to, but don’t make very explicit, is that legibility occurs inside the reader. In this way, it is much like beauty. What we term beautiful things are things that either we personally find beautiful, or things that most viewers will believe are beautiful (out of what subset of observers?). That said, I expect the tips in “How to be Legible” will tend to improve legibility for nearly all readers, so I definitely think it’s coherent to talk about a work as legible in and of itself.
An aside: I initially interpreted “term of art” as something like “a term which is used for artistic flourish”, since I was unaware that it had a definition as a phrase (I thought it was, under my naive definition, a term of art). I was slightly saddened by the fact that it isn’t an industry term (and therefore is not, according to the dictionary, a term of art).
I think the post disagrees with you:
That a work is Epistemically Legible doesn’t mean you’ll comprehend it: you may be lacking necessary background context, for example. See the section Legibility vs Inferential Distance.
In this case, an E-Legible work will still bless you with the awareness that you were missing background context, so that you didn’t understand what was said—as opposed to giving you a fake feeling that you understood. That’s more a property of the work than of the reader, right? You cannot sensibly discount the work’s contribution to such a result.
To be clear, I do believe that the work itself is a significant contributor to the epistemic legibility. My point was merely that the work can only ever be legible with respect to a particular reader or audience (or even expected audience). In this way, I believe it is similar to inferential distance. An idea is not inferentially distant on its own, it is only inferentially distant from someone’s perspective[1]. Where inferential distance deals with the difficulty of understanding a work, epistemic legibility deals with the difficulty of verifying a work. Perhaps an example in which the inferential distance is low but epistemic legibility is low or high depending on the audience will be illuminating:
Rajiv writes a well researched book about encryption algorithms, in which he cites various journal articles published on the subject. Carol is a bright computer scientist with poor judgment who phished people for their bank credentials and got caught. As a result, she is imprisoned and due to the nature of her crime forbidden from accessing the internet while serving her time. She receives a copy of Rajiv’s book through the prison library system, and enjoys it quite a bit. She understands all of the ideas, but wants to check out some of the journal articles to see if Rajiv relayed the ideas correctly and also to further her understanding. The prison librarian checks the price of getting access to these articles, her eyes briefly widen when Carol informs her that she intends to get about a dozen of these articles, and then the librarian tells her there is no way in which she gets any of these articles. Carol is completely unable to verify any of the claims made in Rajiv’s book. A year later, Carol is released from prison, and a couple quick searches reveal the journal articles that Rajiv references, and she is able to verify Rajiv’s claims to her heart’s content. In this example, the inferential distance is low throughout, but the epistemic legibility of Rajiv’s book depends in part on whether or not Carol is imprisoned (unimprisoned Carol could be replaced with some other bright unimprisoned computer scientist).
That was an example of there being differences in the access of information that varies the epistemic legibility of a work. Epistemic legibility might also vary depending on audience when there is the potential for a double illusion of transparency. When this occurs, some audience members will walk away with a confident but incorrect understanding of the work while others might walk away with a confident but correct understanding. This will frequently occur when the language used has variable interpretations (which Elizabeth covered with her discussion of “common knowledge” and inferential distance) or when inferential distance is high.
I don’t think this point is central to Elizabeth’s post (which evaluates works from the perspective of a static reader), so I apologize if I had given the impression that she made some error by not placing emphasis on this. I merely thought it was an interesting frame from which to examine the idea of epistemic legibility.
One point of contrast is that inferential distance likely varies more between audience members than epistemic legibility for a given work. This difference in variability means that it is generally more useful to talk about the epistemic legibility of a work as a whole, but to focus on individual differences for inferential distance
Thanks for the detailed example of what you meant, it’s really helpful. I see what you’re getting at, and it would be insane to say legibility was completely independent of the reader, but I do view epistemic legibility as much more dependent on the work and not the reader than inferential distance, where the load is basically balanced.
In my model, Carol’s ability to check out the articles does not reflect on the book’s epistemic legibility. It definitely affects her ability to check or understand the work, but I’d say the book did its work in being legible by making it possible for Carol to figure out how she needed to follow up, or even just that follow-up was necessary. She’ll understand more when she can check out the articles, but the book met its bar of legibility by not giving her a false impression of understanding more than she did.