Agents Of Fortune

I had one of my thun­der­bolt thoughts after a break­fast of too much cof­fee, and quick­ly went over to write it down: land­scape is a sto­chas­tic art. The form of the design can­not be sure of any out­come, but is fun­da­men­tal­ly open to chance. As an artist in land­scape, you can set out a path and lawns to make cer­tain expe­ri­ences more like­ly, to make cer­tain sto­ries more like­ly. You can incen­tivize peo­ple to slow down and speed up; you can incen­tivize birds and drag­on­flies to land. But you are still far from control.The whole notion of land­scape archi­tec­ture, specif­i­cal­ly, has a vexed rela­tion­ship to this; as I’ve chewed over before, giv­en the land­scape architect’s stan­dard work­flow of gen­er­at­ing instruc­tions for labor­ers to fol­low in a defined con­struc­tion process, the results of land­scape archi­tec­ture priv­i­lege, as much as they can, the sta­ble and the stiff.

But in try­ing to fol­low the exam­ple of musi­cians or painters, who can fall into a rela­tion­ship of con­fi­dence with their mate­ri­als, the land­scape artist finds them­selves at log­ger­heads with their own mate­r­i­al. Every branch on the trees they have ordered is open to the ele­ments and open to haz­ard. A typ­i­cal dis­claimer from a plant sell­er: We do not replace plants that are lost due to neglect, ani­mals, win­ter loss­es, severe weath­er, or plants installed in soil or sun con­di­tions oth­er than those rec­om­mend­ed.” The woe­ful fail­ure rate of street trees high­lights such a lack of cer­tain­ty – no one would put up with struc­tur­al con­crete fail­ing so fre­quent­ly. A dif­fi­cult dia­logue with plants in par­tic­u­lar, pre­dictable at all only as broad aggre­gates and not as sin­gle actors, char­ac­ter­izes writ­ing and thought around gar­dens; it tends to be shunt­ed to the edges of land­scape archi­tec­ture. Plants are will­ful­ly igno­rant of our plans for them.

It turns out that there is a much old­er idea of a sto­chas­tic art, com­ing out of Aris­totelian thought: a sto­chas­tic art is one that can­not be sure of suc­cess, one that can try to per­fect the artist’s aim with­out ever being sure it will hit the mark. It can be for­giv­en if the sto­chas­tic artist fails, because their mate­ri­als are not entire­ly with­in their con­trol. Some typ­i­cal exam­ples giv­en: med­i­cine, gym­nas­tics, nav­i­ga­tion, rhetoric, and, nat­u­ral­ly, agriculture.

The more we think about it, the more we might see the nec­es­sar­i­ly sto­chas­tic aspect of any fine art — the will­ing­ness of any­one dialoging with the art to engage with it on any­thing like its own terms. If, on one hand, the arts of the 20th cen­tu­ry rec­og­nized the dimen­sion of will­ful and way­ward recep­tion, those of the 21st point to the abil­i­ty to quan­ti­fy the like­ly range of opin­ion and inter­pre­ta­tion. That is, if the like­ly pres­ence of fail­ure is one key fea­ture of the sto­chas­tic art, a more pos­i­tive fac­tor today is access to the vol­ume of data nec­es­sary to bet­ter define the rela­tion­ship between pos­si­ble inputs and pos­si­ble out­comes in com­plex sys­tems. This need not mean feel­ing hedged in by the gauge of opti­mal impact. For instance, while dis­cus­sion around Big Data in land­scape archi­tec­ture points toward gen­er­at­ing yet more shav­ing-off of risk or error – rout­ing paths to make opti­mal con­nec­tions, or gaug­ing the opti­mal place­ment of plants – it is just as pos­si­ble to imag­ine the sto­chas­tic as a tool for explor­ing the pos­si­bil­i­ty space of a set of parameters. 

first noise islands
Most are more like Island 37 than Island 33 - but where's the fun in that?
second noise islands
It goes on and on.
error island
Here's Island 60: a very charming error that forgets to make the island at all.

To teach myself Grasshop­per, I have been work­ing on a process for quick­ly gen­er­at­ing unique islands, not unlike the one I made to gen­er­ate forests. These islands basi­cal­ly carve out seg­ments of 3D-mod­eled Per­lin noise – noth­ing advanced for the field. But they help me to under­stand how wide­ly the results of one slate of para­me­ters can range. And the errors gen­er­at­ed along the way are their own sort of treasure. 

It would be a blame­wor­thy thing for a land­scape design­er to spend all of their time build­ing gar­dens on the screen and nev­er putting them into the world. As far as I can tell, aside from pro­toyp­ing Voronoi’d bench­es, the best use for algo­rith­mic tools is in envi­sion­ing pos­si­ble out­comes of more open-end­ed process­es. If such tools seem to have hit a wall in our field as a means to eco­log­i­cal sim­u­la­tion, as far as I can tell it remains to be explored how they can help us envi­sion the results of iter­a­tive form-mak­ing over time. That is, they may be opti­mal for envi­sion­ing a process of land­scape-mak­ing that is exact­ly not archi­tec­ture, but a process, or a folk form, or a game.

(January 2022)