Ok. Linearity isn't my term -- it's jgc's. But I presume he and Carr mean something like a sustained engagement via a body of text. In essay writing, this usually involves multiple revisions, again and again, across the same space. In essay reading, this usually involves a closer level of engagement with a text and its progression than is otherwise normal -- somehow, via its thoroughness, it challenges us.
The disadvantages with this type of approach are that you can get off base. Hence, something more dynamic, with feedback (in the form of a Socratic dialogue, etc.) can be much more useful. Perhaps the Internet has the potential for being more Socratic -- was my point before. But only if we understand exactly how it fits -- how, arguably, we've been waiting for this, in our linearity, for a long time. But if we abuse its dynamic qualities, we might totally lose it. So Nick Carr by reminding us that we might think about these things -- is useful.
The advantages of the linear approach are that by dwelling with something in an engaged manner -- you can often extract value that otherwise superficial readings don't. At some point all ideas come into contact with each other, and are validated, etc. But arguably by giving people more space to develop their ideas -- sort of mentioned in this article:
Your thesis seems rather a reach to me. Socratic dialogues are a sequential form -- earlier points lay the groundwork for later ones, for example.
I'm not an expert on the ancient Greeks, but the last I heard there wasn't much evidence of a written language during the Greek Dark Ages. But to jump from what we know of Greek culture from written remnants (plays and poetry) it would seem to place a great deal of emphasis on "linearity".
Take the Method of Loci used to, ahem, memorize the "linear" points of a speech or argument. While it's true that you can access points at random, that doesn't take away from the fact that they are staged and memorized in a specific order for presentation.
Bottom line, I think you're accepting the Sapir-Whorf hypothesis uncritically. It seems so reasonable, but when you get down to specifics -- like the actual number of words for snow in English and Innuit -- it just doesn't pan out. Similarly, I think anyone who's read Adler and Van Doren's pre-Internet "How to Read a Book" would scoff at Carr's notion that books can't be an interactive medium. People can be mentally lazy using any media.
The Method of Loci sounds very interesting (as well as the Sapir-Whorf hypothesis).
But if we're still presuming that speech is inherently linear, I think we're mistaking how the brain abstracts those sequences of texts. I haven't read the literature at all really, but I read On Intelligence (Jeff Hawkins), and my sense was that this lexing pass that happens when we greet language is happening very much in isolation before we start adding semantic meaning with 'hierarchies' or tree structures, etc. So the 'give off' from that linear feeding of word after word is pretty minimal.
Imho, the Method of Loci would apply greatly to how Greek language and literature works. There is a lot more investment of ideas into spatially separate, almost physical objects (Love is very much a physical god, etc.). This perhaps has the advantage of engaging the participant with those ideas more -- by activating the methods we use in our spatial reasoning. Similarly, in language, the flexibility of the word order really requires that you stack words in different locations, flexibly.
We read the Socratic dialogues linearly -- and Socrates leads us on what seems like a preordained line. But really there are lots of stacks and queues involved. It's a pretty dynamic process of finally finding a linear coherent structure that we're comfortable with. So in my mind it's still very dynamic.
But you make interesting points... Perhaps the best way to evoke the most information with language is to be aware of the distinction between the sequential and dynamic -- and to maximize the good parts of both. They're really yin and yang with each other -- great essay writing is a linear product, made of dynamic error-checking at each step, etc.
The disadvantages with this type of approach are that you can get off base. Hence, something more dynamic, with feedback (in the form of a Socratic dialogue, etc.) can be much more useful. Perhaps the Internet has the potential for being more Socratic -- was my point before. But only if we understand exactly how it fits -- how, arguably, we've been waiting for this, in our linearity, for a long time. But if we abuse its dynamic qualities, we might totally lose it. So Nick Carr by reminding us that we might think about these things -- is useful.
The advantages of the linear approach are that by dwelling with something in an engaged manner -- you can often extract value that otherwise superficial readings don't. At some point all ideas come into contact with each other, and are validated, etc. But arguably by giving people more space to develop their ideas -- sort of mentioned in this article:
http://www.nytimes.com/2010/08/16/technology/16brain.html
-- you extract more value.