In Architecture Depends, author (and dean of architecture at the University of Westminster) Jeremy Till seems to be trying to head off his critics at the pass, by pointing out the book’s shortcomings before they can. After introducing his premise—that architects don’t take into account the unpredictable forces that will change their buildings over time, from weather to dirt to other people’s alterations—he imagines a listener replying, “That’s kind of obvious.” (Till retorts that it may be an obvious point, but it’s still worth writing about, since architects so rarely confront it.)

Yet the real problem with Till’s premise is not that it’s obvious, but that it’s hard to imagine an alternative. How exactly would one plan for unforeseeable changes? Architecture Depends purports to answer, but Till’s idea of an answer is so inchoate and oblique that it’s easy to forget, for pages at a time, what the original question was.

First, it apparently requires redefining time and space. Readers should therefore be prepared for a steady march of sentences like the following: “Space in the Kantian model is a ‘subjective condition of sensibility’ and develops from within the subject so, as Heidegger disparagingly notes, it is as if the Kantian subject ‘emits a space out of itself.’”

The end products of this dizzying metaphysical detour are two concepts which Till believes can lead us towards an architecture that collaborates with entropy instead of resisting it: slack space (space easily adapted to different uses) and thick time (time filtered through human experiences and perceptions). And how should architects employ these concepts? By “infusing the process of design with both experiences and hopes;” by reflecting on the world’s “social and temporal exchanges” during the design process; by “taking the conditions of the everyday into account”… suggestions that are hard to argue with, in the same way that a cloud is hard to wrestle with.

To be fair, there are a few more specific ideas to be found amidst the fog in Architecture Depends, though they’re not necessarily any more actionable. Till laments how architectural representations, from simple sketches to computerized walk-throughs, reinforce the fantasy that buildings exist in an idealized, unchanging setting. Instead, he proposes storytelling—potentially the most productive mode of communication in architectural production, he claims. So instead of informing a client, “You should have your front door here because it is closest to the road,” Till would have architects generate stories about future users’ potential experiences: “We ran through the back door, steaming bodies into air dense with chip fat.”

Setting aside the issue of whether this suggestion is realistic, it still doesn’t give any clearer picture of what kind of building design could actually account for future adaptation, soiling, or weathering. By the end of the book, Till has anticipated this criticism too. “So, I hear you say, what does this… architecture actually look like?” he asks. But his answer is a coy demurral: “If I showed you pictures it would shut down what is meant to be an open argument.” Fair enough. But it would still be nice to have the sense that Till himself has any idea.