jump to navigation

Special/General August 20, 2007

Posted by malthusio in Uncategorized.
trackback

Even now, there seems to be this belief that specialization is a good thing and that this is what complex society is based on. In my mind, this could not be more wrong. If you want to make this claim at all, it should be made in terms of differentiation. The problem with the term “specialization” is that it implies permanence. As McLuhan has said, “Anyone who wants to be taken over by a computer, should just specialize.”. The reason I prefer the term “differentiation” is that it has no implications of permanence. It’s also less purposeful, and to me at least, doesn’t have this built in sense of teleology. Differentiation itself is arbitrary, so it admits the possibility of changing the boundaries.

Of course, it’s also possible that the term “specialization” seems so loaded to me simply because it’s been used in such a loaded way ever since the discipline of history emerged.

The other thing I like better about the term differentiation, is that it isn’t contrary to generalization. So being a differential generalist isn’t a contradiction. Which is a good thing, because this is exactly what society needs. Even where we acknowledge that it is common for people to have multiple careers within a single lifetime (I think the number is at 7 now?), we still look at these from a relatively serial point of view. McLuhan recognized this was wrong even in the 1960’s.

He pointed to the importance of “roles” over “jobs”. The idea being that roles are more all encompassing. They embody a pattern of interaction, not just a single type of interaction, or even a set of specific interactions. You might see the same roles in multiple different fields that are seen as unrelated. This means roles have the potential for stability independent of other changes. Jobs on the other hand are specific. So training for a “job” was like asking to be taken over by a machine. Since machines are better at doing specific things anyway. We’ve yet to be successful in making machines with truly “general” capabilities.

Personally, I’m with McLuhan here. The kind of jobs that machines take over (factory work, manual filing, etc) are likewise not suited to human beings. They are tedious at best and dehumanizing at worst. Anything that can be automated should be automated. This way humans can spend their time doing things they are good at, and things that are better for them. Since the advent of agriculture, people have been doing things that are bad for them, and reducing their general health. Stephen Baxter talks about the consequences that agriculture had on early human civilization in his book “Evolution” (a work of fiction, but you could see where he was going with it), and the tale wasn’t exactly pretty. Perhaps the transition at this stage might be rough, but then so is life saving surgery.

It might seem that automation means giving up control, but this is not the case. The point of automation is to make it so that you no longer have to do the things you didn’t want to do in the first place. Take “watching a movie”, or “getting a massage” for examples. These are things that you would never automate, because that would defeat the purpose. You want to experience these things. For the other side, consider things like “making your heart beat” or “searching the Internet”. Would anyone seriously contend that automating these processes is a bad thing? If your heart didn’t beat automatically you would probably die because you forgot to make it beat while you were sleeping. As for searching the Internet, would anyone seriously want to crawl through sites manually having to guess at what they might find? The Internet would be next to useless.

Generalization is also better for progress in general, because it brings different disciplines together in cross fertilization. There is a combinatorial explosion of possibilities in combining different ideas (allowing exponential growth!). Whereas specialization essentially limits possibilities to a linear function. The only way progress is made is by extending or expanding an existing idea. All paths diverge, but never converge, so ideas can’t be combined. It should be fairly obvious that this is a very limited approach.

Perhaps this bias is partially technologically mediated though. An interesting correlation is that the first technologies were of necessity very general. With pretty much any technology you can see this progression. Take computers for example. The first ones were built out of relatively simple generic parts. There was no other way, because other kinds of parts didn’t exist. Now that restriction doesn’t even apply. Every chip used is application specific (ASIC), because they are so easy to make (even this university can get you an ASIC made if you supply a schematic).

The next thing down the pipe is field programmable gate arrays (FPGA), which can be configured to act like anything you want through software. They are much slower than ASICs, but can be faster if more suitable wired for a task than a given asic. This allows for the possibility of dynamically reconfigurable hardware, which is kind of a neat concept really. When you look at the affect that software running on specific devices, and communication between them has had on society, it’s hard to say what might happen if you introduce general anything devices that are self configuring. While this is a ways off, its not infeasible.

The point I’m making here is that with generalist technology, we seem to have specialist people; but as technology becomes specialized, generalist people are important in order to integrate them. But people adjust more slowly, and going from specialist to generalist is a hard switch – especially when society and education are pushing them in the opposite direction! Not only that, but these phenomena don’t simply move in straight lines. They oscillate, and move in cycles. I’m not sure if I’ve completely characterized these cycles here (probably not), but it does seem to me that they operate along the length of a technological paradigm. With the introduction of a new technological paradigm the technology is general, and so requires specialists to work with it. As the technological paradigm matures, it becomes branched and specialized, and so requires generalists to integrate and apply it.

As you will notice, there are two wave forms here:

  1. The Generalization to specialization cycle of technology.
  2. The specialization to generalization cycle of human users/developers/scientists etc.

They are linked but not synced. If anything, they are at least 1/2 cycle out of phase with each other. The relation here reminds me of current to voltage in an oscillating circuit. The wave form for voltage is always at least 1/2 cycle out of phase with the wave form for current. Of course the situation is more complicated than that, because these two cycles aren’t exactly synced, and it isn’t clear to me exactly how they are linked. I am fairly sure that they are linked though.

Advertisements

Comments»

No comments yet — be the first.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: