When Tech Turns Nouns Into Smarter Nouns

Of late, the New York Times’s Bits blog has proven a bona fide wellspring of linguistic thought experiments, if only accidentally. Not long ago, the technology blog prompted me to muse about the mechanics of a language composed solely of pictures. This Monday I stumbled across a post by Quentin Hardy beguilingly entitled “When Tech Turns Nouns Into Verbs”:

“The phone is a little connected computer—a device whose uses and meaning we continually explore and modify. It is by no means a phone in the historical sense. It is still a physical object, of course, but it is really a vehicle for one or another software-enabled experience. In an important sense, it is made to be contingent, changing with every download and update. That focus on the needs-driven experience means it behaves less like a static noun and more like an active verb.”

Most of the objects in our lives, Hardy goes on to say, are becoming smarter, increasingly able to learn about us, adapt to our needs, and even spy on us. Well, okay.

But Hardy loses me—and yes, I’m being deliberately obtuse here—at the notion that these objects will soon behave less like static nouns and more like active verbs. Were I to be either a noun or a verb for Halloween, I wouldn’t know where to begin. The problem is that many verbs aren’t active in the least: in English, we exert and go, but we also rest and stay; more often still, we are and could have been. Nor are nouns particularly static. I’m a noun. So are trains and explosions and Swiss Army knives complete with can openers and corkscrews. I know. The point is simply that objects will become less static and more active.

And yet, the underlying linguistic question here is pretty interesting. If the objects around us do become considerably more lifelike—if, as Hardy put it, “intelligence is everywhere”—surely this would have some impact on the language we use to describe our interactions with the world. We’re highly unlikely to stop slapping nouns on these newly intellectualized objects. But we may indeed use these labels in ways that betray their increased abilities.

I’m reminded of semanticist David Dowty’s work on “prototypical” agents and patients, terms bestowed upon nouns in particular relationships with the verbs in a sentence. Roughly speaking, in Dowty’s framework, Proto-Agents are in control of their own destiny: sentient, mobile beings able to deliberately act on and cause change to their environment. Proto-Patients, on the other hand, are insentient and immobile. They are not movers or shakers, but rather the entity moved or shaken.

Consider an event described by a transitive verb: let’s say a collecting event. The collector is an agent, as it has more of the properties of a Proto-Agent, and the collectee is a patient, as it has more of the properties of a Proto-Patient. This is true even if the specific nouns that fill these collector-collectee roles do not themselves have all of the properties of Proto-Agents or Proto-Patients. Thus, inanimate databases can collect information, and we can collect lively friends. Still, it’s probably safe to say that nouns with Proto-Agent properties are likelier to be agents.

In active (as opposed to passive) sentences in English, we put agents in subject position. This gives us certain expectations about the types of nouns that are likely to appear as subjects versus objects. To see what I mean, take a nonsense sentence like The rom mecked the zarg. If you are like most people, even without having any idea what mecking is, you’ll intuit that roms are more likely to be aware, mobile, willful, and capable of effecting change than zargs. You may be wrong—after all, you aren’t inherently more sentient or proactive than your friends. But on the whole, the rule will serve you well.

Now consider an interaction between Kevin and a thermostat. In this event—a temperature-adjusting event—Kevin is very much a Proto-agent, and the thermostat a Proto-patient. We would probably describe this scenario using a sentence like Kevin adjusted the thermostat.

But what if the thermostat in question is a very smart thermostat—a thermostat that anticipates Kevin’s needs based on past heating and air conditioning usage, one that perhaps even detects that he is preparing for bed early, or that company has arrived, and switches temperatures accordingly? Now it is the thermostat that, when it comes to adjusting the temperature, seems to play a more agentive role. We might describe this scenario as The thermostat adjusted or even The thermostat accommodated Kevin.

With intelligence distributed more widely among the objects that surround us, in other words, our own role as an agent—speaking linguistically and, I suppose, non-linguistically—might shift as well.

How else might a proliferation of smarter, more agentive nouns lead to language change? I’d love to hear your thoughts. In the meantime, keep the language posts coming, Bits.