A brief demonstration

I was wandering around Harvard the other night, mostly in celebration of being ambulatory enough to leave the house again. (Dragging myself to CVS like a snot-ridden zombie doesn't count.) I do that, from time to time, especially when I'm going stir-crazy -- after growing up in Phoenix, I find it very soothing to walk around where four hundred years of university have soaked into the ground.

I live about three miles off. It was dark by the time I got to the Yard, and I could see flashing lights through the trees. Three fire trucks were pulled up outside one of the buildings. I finished my loop of the green and wandered off down Mass Ave again.

It wasn't until I was walking away that I realized that this is probably the sort of sherlocky thing other people are curious about, and I might as well write it up.

By far the largest part of this superpower is just paying some fucking attention. Not walking smack into ladder trucks is a great start. The trucks themselves weren't in view from outside of the Yard proper, but the flashing lights they cast across the buildings peeked out at the corners, visible from the old church on the corner, on the other side of the Cambridge St traffic tunnel. The lights on emergency vehicles flash in a regular pattern; the flickers were off-kilter, so there was obviously more than one in attendance.

The second largest part of the superpower is asking questions. Shotgunning questions at random doesn't work either; there's a lot of information in the outside world, but you need to do some gatekeeping in order to get anything useful out of it. All the information in the universe does you no good if you have no way to sift through and match pieces up to make it knowledge.

So, to start with: In what situations would one call for the kinds of emergency vehicles with flashing lights?

Narrowing it down, when I actually got onto the green, I could see that there were two ladder trucks and an auxiliary. I haven't got the entire map of the Cambridge fire districting stuffed in my head, but there's a station  less than a mile away, down on Cambridge and Kirkland, which would probably actually be visible from the inner edge of the Yard if not for terrain. Also notable is that they were all Cambridge FD trucks. EMS in Cambridge is based in fire stations and a lot of firefighters are cross-trained as paramedics, but EMS is a private company and FD is run by the city; for uncomplicated medical calls they tend to send out a proper ambulance and skip the ladders. No cops -- Harvard has its own PD and there are Cambridge police operating in the nearby Square -- so no immediate dire hazard. Fire alarm or hazmat spill, then.

Narrow it down: How could you differentiate between these two occurrences without being there when it happened, talking to the firefighters, or being allowed inside?

There was no crowd around the building, that I could see. Also no formal perimeter. A couple of firefighters were standing next to one of the trucks, talking. This in itself tells me nothing: No people could mean no one was evacuated, or just that no one much was inside when whatever it was happened. There's also the possibility that everyone was shooed out and went home, but it's unlikely on a college campus. Classroom buildings are evacuated very quickly in emergencies, enough so that people are forced to leave a lot of things behind, and students can get very antsy about missing class time if attendance is graded. If it were a block of offices, people still there at 7:45pm are people who are working on something, and have had to leave it behind.

All right then. What's inside the building and would anyone have been in there with it?

I don't know the names of anything surrounding the Yard and Harvard is epic fail at signage -- their attitude is that the place has been there for four centuries now and if you still can't find your way around it's your own damn fault -- but it was well after class hours and there were lights all over the place around the green. Given the plethora of eclectic and mismatched crap visible through everyone's open blinds, I'd guess all of those buildings are dorms. (Supported by my discovery last year that you can actually park the car at Harvard Yard, but only during freshman move-in weekend.) Fire engines in front of a dorm but no crowd outside means that everyone's already been let back in. The FD doesn't let people back in if there's a known fire or a hazmat problem; the dorm staff would let people back in fairly quickly after a drill, but wouldn't have called for ladder trucks.

Another significant part of sherlocking is pattern matching. This, I don't know how to teach -- you just have to look at combinations of things and ask yourself, where have I seen this before? If it doesn't come to you, it doesn't. If it does, it's easy.

Lights on + crap visible through the windows = dormitory.
Three fire engines = real alarm was set off.
Real alarm + no crowd + no perimeter = no building hazard.

How do you set an alarm off in a dormitory without actually setting anything on fire? Well, the most popular option is to accidentally throw your popcorn in the microwave for thirty-three minutes instead of three minutes and thirty seconds, and then walk away. (Second-most popular is to forget to cover the smoke detector before trying to hotbox your dorm room. Cops are normally involved in that, though.) Nice to know the Harvard kids can't cook unsupervised either.

The last major part of the sherlocking is, of course, showmanship. Since I had no one with me to show off to at the time, I just calmly kept on my walk without breaking stride, having figured out nothing of any real interest was going on with two or three glances around, in about fifteen seconds.

It's fairly difficult for me to break it down well enough to write it out for others, because it's so quick and I do it so constantly. It takes significantly longer to explain than to actually do. Each piece of information brings up its own handful of different applicable contexts, and they sort of rotate against each other until a bunch of things match up on all the different gears, and that's the answer. You know that a pan on a stove with food sizzling in it is probably hot, I know that a bunch of bored firemen standing around their trucks alone in front of a college dorm means some idiot fat-fingered the timer on his snack. It's a more complicated set of indexicals, but that's all they are -- outward indications of something else happening or having already happened where I happen to not directly see.

The more complicated things with people in have to do with correlating behaviors with mental states and motivations. That is something you can do without ever really developing empathy for the reasons people do things, if you're sharp enough to catch patterns in the abstract, but empathy makes it much easier by introducing the element of fuzzy matching. If you empathize with the motivation, or at least intellectually comprehend the emotion behind it, it gives you a better idea of what that person is going to try to express in their actions, and hence what kinds of actions you might want to be on the lookout for. As a side effect, you start being able to explain people more eloquently than they can explain themselves -- most people don't really spend much time trying to puzzle out how they work internally, and often have absolutely no idea why on Earth they do the things they do. You generally don't get the exact history behind things like this; you don't know, for example, whether someone grew up in an abusive home or was abandoned or has gone through a very painful breakup or a death in the family or some sort of separation, but you can come to recognize quite well when someone's acting out of a fear of abandonment.


  1. Pretty sure what you're describing is just fluid reasoning. But what I think is interesting is that while fluid reasoning is correlated with IQ (especially in tests like the WJ where it's a subtest), it's really not as strong a correlation as we tend to think. But so far, fluid reasoning has seemed impossible to teach, while a number of testing subtests are eminently teachable (all of the crystallized knowledge tasks, some of the processing tasks).

    1. It might be; I'm not surprised it has a name. Mostly I know that when I do it, people gape at me in amazement for some reason. The process is probably universal or nearly so -- it seems a basic human survival skill -- but part of the reason I seem to be very, very good at it is that I can also bring into play enormous volumes of otherwise unconnected facts, observations, and context. This is just a simple example that I could take apart well enough to write out; it can get horribly complicated very quickly when the situation gets more elaborate and involves people, singly or in groups. Explaining those coherently is nearly impossible. Language is linear, and the map of stuff I'm snapping together is of a much higher order in my head, with no good 2D path to follow. I'm either backtracking and repeating myself constantly or spitting out a lot of what sound like completely unrelated bits and pieces to people who aren't in my head.

      I always thought the Holmes thing was part native talent and part skill. Other people don't seem to. I might just see it that way because I have both pieces.


Post a Comment