(This is the first of what will hopefully be a series of posts on using digital tools to think through the structure of sonnet sequences.)
This is the edition of Sidney’s Astrophil and Stella I’ve been (ab)using since I was an undergraduate:
At this point, it’s less book than folder, holding sheafs of poems with several layers of annotations. On the right hand of the page, most of the notes attempt to wrestle with this one particular poem. Glossing puns, marking images, and tracing connections, these notes track the efflorescent activity of sense-making. Reading through this edition, I can see myself learning to read poems: I scan lines, mark rhyme schemes, hunt for voltas, and so forth.
On the left is remnants of another type of activity: trying to track these same clusters and themes through the poems. We fundamentally misunderstand the sonnet sequence, I would argue, when we try to read it as a series of particular poems. Rather, meaning accrues through the repeated recombination of significant phrases and images. An image, posture, gesture, or position will be mentioned in one poem, developed in the next, and henceforth alluded to with the full force of context. Repetition with variation becomes architectural. But how do we analyze such structures? How do we even represent them to ourselves so as to track them more fully?
My right page tries a couple of strategies: a list of references (to the “horse” cluster) in the bottom left; at top an effort to follow two terms that slowly disintegrate through the sequence. I’ve always struggled to write about these effects because they’re subtle, multiple, aggregative, and contextual. (I sketch out one such cluster below.)
For about as long as I’ve been annotating that volume, I’ve wondered whether hypertext will help us to model and reverse-engineer these types of connection. Hypertext lets one follow connections between poems other than those of sequence. So, over the past couple of weeks, I’ve been experimenting with modeling Astrophil and Stella in Twine 2.0. Twine is a terrifically easy to use tool for building hypertext stories: it’s frequently used to create text-driven games.
Here, I’ve built out the sequence as a collection of linked pieces of text. I’ve chosen to make each poem or song have its own page. (On Twitter, Claude Willan suggested that it might be equally interesting to give each quire a page.) I’ve pasted the text of each poem on its page, using the text from A. S. Kline’s open modernized text.
I’ve then started to connect key phrases to later poems that develop them. You can see here the form of the links: two brackets surrounding the word that will be linked, a | character, and then the passage linked to. “[[force of heav’nly beams|36]]” links that phrase to sonnet 36. And here’s the same poem when I run my Choose-Your-Own-Edition Sidney.
I am slowly putting together an idiosyncratic personal edition that lets me track my experience of particular image clusters in Astrophil and Stella.
As you can imagine, this is slow, associative, and imprecise work. Here’s an in-process shot, as I work my way through the first few poems. I’ve got a lot of work left to do, but I’m already getting a new sense for how this sequence builds meaning.
Even at this early stage, I can tell that Twine isn’t ideally suited for modeling the types of relation I’m interested in. Hyperlinks connect two particular moments, unidirectionally. But how do we decide which two moments to connect?
Take one particular image-pattern as an example: the forceful beams from Stella’s eyes. The first example appears in the catalog in 6: one thing lovers write of, Astrophil says, is the “force of heavenly beams.” Sure enough, 7 takes on the challenge directly:
When Nature made her chief work, Stella’s eyes,
In colour black why wrapp’d she beams so bright?
Then, 8 riffs on the idea: what Cupid enjoys living in Stella’s face is her “beamy eyes.”
So these poems form a cluster, with both 6 and 8 alluding to an idea that is worked out at length between them. Once this set of associations is built, Sidney can draw on it through a brief allusion:
Long since, forc’d by thy beams, but stone nor tree,
By Sences priviledge, can scape from thee!
Stella lookt on, and from her heau’nly face
Sent forth the beames which made so faire my race
And in turn, these two passing allusions set up a more straightforward reassertion of the image in 42:
O eyes, which do the spheres of beauty moue;
Whose beames be ioyes, whose ioyes all vertues be…
And, in turn, now that the image cluster has been laid out, it can change in tone:
What, haue I thus betray’d my libertie?
Can those blacke beames such burning markes engraue
In my free side, or am I borne a slaue,
We see the forcefullness of “beams” as in 6 and 36 paired once more with the blackness of Stella’s eyes, now as a branding iron (also black, but glowing when heated). And immediately thereafter, in 48, this turn is consolidated as Cupid’s “beamy dart.”
Both “dart” and “slave” here are attached to image clusters of their own, tracing their own lines through the sequence. (If from here we read backwards, we find “O eyes, dart down your rays” as foreshadowing back in 42.)
This is just one image, followed halfway through the sequence, but it’s already exposing problems with my modeling. How do I decide where to link the relevant phrases? Do I connect to the next poem to involve the cluster, emphasizing the sequentiality of these moments? Or do I connect each instance to the nearmost poem that really develops this image, in more of a hub-and-spoke model? Once “beam” and “dart” combine at 48, should I link backwards to poems that relate these ideas less directly, like 42? Or forward to moments which continue to develop one of the two? While good at coordinating parallel moments, hyperlinks don’t quite do justice to the multiplicity of connections.
In future posts, I’m going to be experimenting with other digital ways of modeling these connections. I’d appreciate any ideas you have.