Roots and Leaves

Recently I read Poetic Meter and Poetic Form by the essayist and scholar Paul Fussell. As the title implies, the book deals with the nature of meter and form in (English) poetry.

In high school English class there was a unit on poetry where we would learn about the same technical elements that we did in the previous year: meter, caesura, enjambment, and the others. The only thing that changed from year to year was the selection of poetry we analyzed: sometimes I really enjoyed the selection (T.S. Eliot) and other times I detested it (Billy Collins). No matter what we read, analyzing the poetry often felt hand-wavy and dishonest, because we weren't taught much about poetic techniques besides how to recognize them. That is, we would be told what enjambment is, but never why an author uses it or what effect it has on a poem.

I bring this up because Fussell's book has answered a lot of the questions I had back then, and I finally feel like I can begin to analyze poetry in earnest.


When I was on vacation I had the opportunity to visit an apartment in Venice where Ezra Pound lived for awhile with the concert violinist Olga Rudge, who was his mistress. The apartment is in a relatively quiet alleyway in the eastern part of Dorsoduro--one of the 6 sestieri (~subdivisions) of Venice. Above the door was a big plaque memorializing Ezra Pound, a TITANO DELLA POESIA. The doorbell was still labelled in permanent marker with Rudge's name and address. Pound and Rudge are both deceased, so I wonder who owns the place these days, and what it's like inside.

I also went to a museum of Tyrolean culture and was fascinated by the depth of symbolism, and level of craftsmanship, that many of the artifacts possessed. It must take a long time, and relative isolation, for such a distinctive culture to form. Even looking at something as simple as a calendar, you could see how a deep and primordial sense of mysticism must have permeated medieval life.

In contrast, the modern culture of central Europe--based on my experience--seems really bourgeois. Obviously my writing style on this blog is bourgeois and affected, which is a habit I need to break: I want to be a creator and not a scholar. Scholars use the bourgeois style of prose to signal their intelligence and sophistication, while any creator worth his salt should express things in an authentic and unencumbered way. So while I may be pretentious, I prefer the rough but friendly culture of America to the haughty formality of central Europe. Particularly in the German-speaking countries I felt like I was in a public library--though part of this feeling came from self-consciousness, because as a pseudointellectual I didn't want to appear to the locals as just another poorly behaved American.


On an altogether different note, think of the natural beauty that a tree has. People have devised procedural generation techniques that can sort of replicate the form of a tree, but not quite: notable examples include L-systems, fractals, and differential growth algorithms. In addition to this, machine learning models can analyze the shape of a bunch of trees and then generate lookalikes that are based on quite inscrutable parameters that the model "learned".

What I would like to know is whether a model exists that replicates the shape of trees in a realistic way that humans can understand. L-systems and fractals are simple to understand but unrealistic, while the inverse holds for ML models.

Trees, in all their diversity, grow in a way that one perceives as "natural", and the hallmarks of nature resonate with our sense of beauty: which is why, I would argue, the work of someone like van Gogh can seem so un-realistic yet appear so real to us: natural proportions appear in the abstract, giving rise to unexpected beauty.