Mario Carpo


Mario Carpo is the Reyner Banham Professor of Architectural History at the Bartlett School of Architecture

interview by correspondence amsterdam - london, - January 2019

“Big data computation, simulation, optimization, and machine learning...!” - Mario Carpo


Ayax Abreu Garcia: Review Architecture tries to make architecture theory alive and useful. Given the variety of books there are in the profession, allowing there to be not only strict 'theory' texts but also books about construction, phenomenology, monographs, etcetera. In your opinion, what is the most influential architecture theory book and why? In which ways would the profession be different without it?

Mario Carpo: Speaking as a historian of architectural theory, in pure quantitative terms, and regardless of my personal inclinations, I would say first ex-aequo Vitruvius and Alberti; then Corb, Vers une architecture 

AAG: What is your research methodology and how do you apply it when writing a book? Do you read a book and follow the lead wherever it takes you, or do you start with a specific reading list?

MC: I regret I do not have any methodology. When I have something to say, I look for arguments to corroborate my intuitions; if I find them, I write them down. That’s it.


AAG: What was the inception or motivating factor for writing the Second Digital Turn?

MC: I had a question: why is the architecture of the second digital turn so different from that of the first? why is it that twenty years ago all my friends made splines, or continuous curves and surfaces, and today they all make discrete, particlized volumes and dissonant, chunky, messy stuff? I am satisfied that I found a believable answer--until evidence of the opposite--so I wrote it down.

AAG: As technology keeps progressing, what will be the role of architects in the future? Will all architects have to be conversant in computational design?

MC: Pretty much so, to some extent.

AAG:I really enjoyed your book, I was even laughing at some points, like with the anecdote of your professor that never showed up again leaving you without an answer to the arithmetic problem, I think the anecdotes are visualizing narrative tools for the reader to comprehend your points, plus they make the book more enjoyable.

Were the anecdotes in the book there for this purpose?

MC: Yes. I do not know if anecdotes make the book more enjoyable--that depends on the inclinations of the reader; but, ideally, an anecdote is a silly story that conceals and conveys some deeper meaning. We remember stories more easily than we remember abstract concepts, so storytelling is a common mnemonic device. Unless one is a mathematician, of course... mathematicians use equations. That’s shorter and more effective. But I am not a mathematician, so I don't.

AAG:Additionally, I really liked the historical examples that you provide throughout, such as in the end of the projected image chapter with Alberti's Measuring device for statuary replication or Henry Dexter's Apparatus for Sculptures to be employed in copying busts, to me those examples show the ingenuity of artists and how, in pre-modern days they thought about computational problems. It feels that you grounded contemporary themes on a larger conversation. 4. Why is finding historical precedents so important?

MC: Because it saves time. We do not need to reinvent the wheel every time anew--but we may do so if we do not know it has been invented already. See a recent short piece I published on arch-daily and metropolis, "a plea for architectural history".


AAG: Are the examples on the book that display what you call 'excessive resolution' aesthetic archetypes? In other words, will their aesthetic be recognized as the aesthetic of the 2020s and 30s? Or is it just the so-called avant-garde? (i.e. Achim Menges, Michael Hansmeyer, Daniel Widrig)

MC: I called it "excessive resolution", but also discretization, voxellization, particlization, etc.; a forthcoming issue of AD, edited by Gilles Retsin, offers a palette of similar terms. I think AI and robotic assembly are confirming that trend, but also already suggesting some new developments.

AAG: In the chapter The participatory turn that never was you said "Particularly in the digital model of open-ended aggregation, the effectiveness of the result is achieved not by dint of authorial precision, but through approximation, redundancy, and endless participatory revisions" You could argue that the act of creating architecture now at days is already quite collaborative, but do you see a future where a project keeps being revised as a wikipedia entry?

MC: Architecture (in the western, humanistic, Albertian tradition) is an authorial art par excellence. It was born as an authorial art an art of design and notation, and that is unlikely to change anytime soon. That’s one reason why, as I argue in that chapter, "the participatory turn" in architecture has not happened so far. But technology is now going in a different direction, and in cultures less influenced by the western humanistic tradition (in parts of Asia, for example) the loss of the authorial status of the modern western designer may not be as problematic as it is bound to be in Ivy league universities. 

AAG: You talked about Pierre Bezier and Paul de Casteljau on their splines research as "appearing to have been primarily motivated by the mathematical ambition to translate general free-form curves into equations" and therefore opening a whole new era for digital architecture and design. In your opinion, what is the contemporary version of this paradigm-shifting event?

MC: Big data computation, simulation, optimization, and machine learning...!

AAG: What interests you in knowledge that could inspire people to keep theory more alive?

MC: No need to keep theory more alive than it should. What drives me is the will to understand what the heck is going on. All means are good to that end...

further reading: the second digital turn