It’s one of my favorite parts of a repair job: a symbolic transition, instantiated in the physical act of switching the ratchet from “loosen” to “tighten.” In terms of the sense of satisfaction it brings, it is rivaled only by the moment a newly-repaired mechanism starts and operates properly again for the first time. For most of the ratchets I have owned, it involves rotating the dial counter-clockwise on the back of the head. On others, there is a lever easily flipped from left to right with a quick motion of the thumb. Either way there is a tangible sensation, even if the “click” is not audible, that signals a job well done.
I was 11 or 12 years old, working on a small engine (probably my lawn mower, though it could have been my gokart) in the garage of my family’s home, in Eden Prairie, when I first recognized the click was a symbolic-kinesthetic marker of the transition to reassembly. I’m not sure why that memory is so clear. My sesquipedalian topic sentence belies the fact that it was such a mundane realization – too obvious to really be an insight. Still, it seemed like an epiphany to me at the time: the click marks the halfway point of the repair job. It happens because you have finished disassembling and are about to begin reassembling – all of the same steps, just in reverse.
I’ll bet my friend Paul was busy with CCD or something, and in the space created by the absence of conversation, I could be reflective.
Or perhaps the original memory was cemented by years of recollection – and the growing realization I had been wrong. My memory of this, later realization is far less clear though. I can’t tell you where I was – or even when I was. But at some point, it struck me that reassembly never takes as long as disassembly. I have reflected on it, and reasons for it, dozens of times over the years (since my family moved from Eden Prairie, Paul hasn’t been around to fill the conversational space, so I get a lot of thinking done with grease on my hands). At first, I thought it was just that a whole bunch of cleaning accompanies disassembly. Then, that a whole bunch of learning accompanies disassembly. However, while it is true that dirt, grease, and mystery are removed by the time I start putting things back together, I’ve come to see a different reason that it always goes back together faster: reassembly is merely a formality.
If you are paying attention to what you are doing, the whole of the repair is accomplished as part of disassembly. That is when all of the forensic work is done – and that is the intensely intellectual part of the task. In disassembly, you discover how all of the parts fit together – and in so doing, construct a mental model of how they work together. Even if you know, a priori, how a thing works, disassembly is always a chance to refine that mental model through the forensic work of checking specific parts for telltale wear or damage. It is on that mental model that the real work of the mechanic is done. When you know how a thing is supposed to work, you can begin to imagine the consequences of particular damage or failures of specific parts – and search for one that matches the actual symptoms that precipitated the repair job. It is a forensic task.
If describing it that way makes it sound fairly straightforward, I should clarify: it isn’t.
Most symptoms match numerous potential causes – and working out how, and in what order, to test competing hypotheses about possible causes is itself a complex logistical problem. Sometimes the least likely cause gets investigated first – if it is inexpensive and requires little effort to fix. I have often made a simple repair, even though I didn’t really expect it to fix the problem, on the grounds that I may as well take 5 minutes to rule it out before investing hours into addressing the more likely cause. The worst that can happen in that case is that I confirm the unlikely thing was not the problem (and ensure it continues not to be a problem). On the other hand, if the unlikely thing does turn out to have been the problem, well then I’ve saved myself a bunch of time and energy (and probably expense) unnecessarily fixing the more complicated thing.
I would like to say I developed this common sense rationally, but the truth is that I had to make the mistake (more than once) of diving in to fix the complicated thing first – and feeling foolish after discovering the solution was much easier. Feeling foolish is instructive. (Well, it can be. If you let it.) Experience has taught me to be certain the sound of hoofbeats is not a horse, before I start looking for the zebra.
[note: On the occasions that I have invested a bunch of time and energy into unnecessarily fixing a complicated thing, a result has been a new and deeper understanding of how the mechanism works. So it’s never really wasted time to do that. But there are only so many hours in the day…]
It is the rare case that a root cause of malfunction is due to anything that is obviously broken. Unless something has gone catastrophically wrong, it is far more common for primary causes to be too subtle to notice unless you are specifically looking for them. Matthew Crawford describes a ‘scales-falling-from-the-eyes’ moment in which a more experienced mechanic pointed out to him subtle wear patterns on valve-stems from his engine, that indicated someone had installed stiffer than specified valve springs years earlier. What had been completely invisible to him moments earlier was suddenly so obvious he couldn’t not see it anymore. I had my own, similar experience at the hands of a more experienced mechanic. It was at the first small engine shop in which I worked. He showed me a float-valve needle and asked if I thought it needed to be replaced. After I said it looked fine to me, he pointing to the faint ring completely encircling the tip, which showed I was wrong. For both me and Crawford, what needed to be done to affect the repair became obvious – once we suddenly saw the problem. But problem finding is the hard part. Learning to see subtle evidence like this takes experience. You have to know what you are looking for – and you have to be intentional about looking for it. You need a hypothesis to test. And that is at the core of the craft of mechanicing. It is how mechanics do the problem-finding part.
It seems noteworthy to me that information provided in repair manuals is really only of value to the already well-informed. If you are working on an unfamiliar machine or assembly, repair manuals often contain helpful tips for avoiding damage to internal parts, working within confined spaces (e.g., engine compartments or wheel-wells), making sensitive adjustments, etc. But books can’t troubleshoot. They have no capacity for thought, action, or skill (more succinctly: books, like telephones, are stupid). And what I am trying to illustrate here is that troubleshooting absolutely requires all three: thought, action, and skill. Thus, you need to already know what needs fixing in order to use a repair manual. And to know what needs fixing, you either need to already have a well-formed mental model of the thing to be repaired – or get advice (or help) from someone who does. Repair manuals are therefore, when used properly, a good resource for ongoing development of the mechanic’s craft. However, when they are used mindlessly, they pose the danger of short-circuiting the critical thinking essential to developing craft knowing. In any case, I have come to view reassembly instructions in repair manuals as a useful audit system to ensure I don’t forget steps – but a poor source of guidance.
Over nearly four decades of breaking stuff, and almost as many years of fixing stuff, I have developed the habit of noticing details during disassembly that might be important during reassembly.
Hmmmm… said that way, it sounds passive and haphazard. Let me try that again…
The habits of mind of a mechanic are to actively seek data during every part of a repair and to analyze those data through mental simulations of relations to the whole of the mechanism and its function. As I wrestle parts out of awkward spaces, I note orientations that need to be incorporated into reassembly strategies – and make decisions about reassembly order or location to avoid awkwardness as much as possible (especially when sealants or delicate parts are involved). My fingers notice pressure and tightness of each nut, bolt, screw, belt, spring, lever, etc. And I start strategizing about which tools to select for reassembly (this can be especially important when springs are involved – getting the right combination of clamping pressure and tool orientation can be a real PITA for installing heavily loaded springs in tight spaces, like drum breaks). And I look for parts that can be repaired or replaced proactively, even if wear or damage is not yet visible to me, so that I don’t find myself back inside this assembly again when a different part fails. These are all outgrowths of having made decades of mistakes – and learned what to avoid by experience.
So, by the time I start putting stuff back together again, I have an elaborated mental model of how the mechanism works, a strategy in mind for reassembly, a list of potential mistakes to avoid, and a clear image of what it will look like when everything is reconstituted. In other words, the click signals that the repair job is done. All that remains is the formality of putting it all back together – and to use.
>>>>>>>>>>>>>>>>>>>>>>>>>>>
Coda:
The hollow thump of the regulator switch on an impact wrench evokes a different feeling for me even though it serves exactly the same purpose of switching from loosen to tighten. Jobs that can reasonably be done with that much power are usually accompanied by copious rust, dirt, and profanity. So their denouement is usually relief rather than satisfaction. Then again, I have less experience with this type of repair job. Perhaps I just don’t have enough perspective to appreciate the beginning of the end in those cases. But perspective is hard to find with dirt and rust falling into your eyes and mixing with the blood welling from mangled knuckles.