Considering broader details than specific cases in isolation

So many people are only able to consider specific cases (like something controversial that a person said) in terms of the details of that case, and that considered in isolation. But for so many topics, it's essential to be able to consider broader details that apply to the specific case (like free-speech). There are many many examples of this, across all sorts of different domains.

In part, this post is for me to clarify my understanding of this topic and to improve my ability to express my thoughts on it. It could be refined a lot, but my main priority at the moment is to just get something out there that's a starting point.

There are cases where someone has said something that some people agree with and think is good, while other people disagree with it and think it's bad (and perhaps even harmful enough that they think the person shouldn't be allowed to say such things).  Some people can only consider such cases  in terms of the specific details of the case.

They might see that utterance as wrong, and something that shouldn't be said. And they can't go beyond this to broader principles, that might take precedence here, such as free-speech.

Not only that, but for many people they can only see the particular cases in terms of their own point of view of those cases. They take their point of view on it for granted, as if it was a basic fact about that case. So say that they disagree with what was said. They believe that the case should be handled from the standpoint that what was said was wrong.

Many broader principles can't take a stance on whether particular views are right or wrong. They're above that level. Like free speech. Because if topics are disputed, who gets to simply declare what is the correct position? Whoever could do that would have far too much power. 

But these people can not get past their perception that the view in this particular case is wrong (or right). To them there's the issue of free speech shouldn't even come into consideration in this case, because the speech was wrong.

Because many people only view things in terms of the particular cases, they interpret recourse to principles in those terms as well. So say that in a particular case there is dispute about which position is right and which is wrong. They see that case in such terms, and they are taking a particular side. If someone tries to invoke a principle, like free speech or the rule of law, as being applicable in this situation, then who this person sees as the guilty party may, for example, be allowed to say the bad thing. Thus, invoking the principle is interpreted as taking a particular side in this particular case. Anyone who has tried to invoke such principles will likely be familiar with this sort of response.

Because they can only see it in terms of this particular case, if this case seems unimportant or trivial to them, but you're making a fuss relating a principle like free speech or the principles of fair argumentation, it'll seem to them that you're making a fuss over nothing. They won't get that it's about principles, that need to be defended in general. An analogy: if someone drops a chip packet on the ground and a nearby police officer calls them out, that person might respond that it's no big deal, it's just one chip packet and that's not going to make a difference. But they're focusing on that specific situation. Any one case considered in isolation doesn't seem a big deal. If every person dropped their "just one chip packet" on the ground, these broader details of dropping chip packets in general are a big deal. The police officer can't consider every case in isolation, they have to think about the general principle.

Other examples of such broader principles include due process, the rule of law, and principles of fair dealing people and of fair argumentation.

I mentioned earlier how people often can't get past their own particular beliefs about a matter, effectively treating them as something that everyone else should take for granted. This kind of thing happens in conversations and arguments as well. Where people can't past the belief that their view X is correct. They can't see things in terms of broader principles of fair argumentation. That you can't expect other people to accept the truth of your position on your say so (for that would require you do so for their position). That you have to argue why it is true. Too often people take the truth of their position as a starting point for how they think the argument should be conducted.

Being able to rise above particular cases means being able to handle more abstraction.

It's important because it's important for understanding systemic issues, and so many important societal issues are systemic issues.

At this point I can talk a bit about what broader details actually are, and what sorts of things are missed if you focus on specific cases.

If you just focus on specific cases, you're never seeing trends or patterns. You just see case one, then case two, then case three, and you never consider case one AND case two AND case three, and what that totality tells you. You don't see cumulative effects of all the cases. You don't see the effects over different time scales.

And it's not just the instances of these kinds of things over longer time frames, it's how the instances of these kinds of things play a role in larger systems over those timeframes. In the case of free-speech, it concerns how speech across a population interacts with political happenings and power structures. How entities will try to control speech to consolidate their power.

So far I've been talking about being able to look past specific cases to broader principles that apply. There's named principles, like free speech and the rule of law. But there are other kinds of broader details that aren't really principles (and don't have names) that you miss if you only focus on specific cases.

The broader details are important in technological development. If we look at technological development through the lens of specific cases of new technological developments we will greatly misunderstand the nature of technological development.

Any individual development can seem fairly trivial and unimportant. Say we consider some new development in video games. To many people that's just a development in video games. Just something in entertainment, and unrelated to "the real world".

But the broader details here include that technology tends to be fairly 'content neutral' and that technology builds on other technology. In terms of it being 'content neutral', something developed for a particular application can often be used or built on for other applications in other domains.

And technology builds on other technology. One small improvement in gaming might be something that can be built on in some other domain and used there.

And from a broader perspective, we don't just consider individual cases in isolation, but we look at those cases together. A number of "minor" developments over the years may together enable much larger and more significant developments.

There are other broader details, like that the greater the amount of computing hardware that is produced, the cheaper computing hardware will be in general. So things that increase the use of computing, like gaming, also contribute to making computing cheaper and more accessible in other areas.

Some kinds of technologies have an impact on several fields. Consider information technology. Information technology is a central part of most fields these days, so when we consider the impact of new developments in information technology, the broader details include the impact that will have within all of those other fields.

One thing to appreciate about information technology is that a lot of it is "content free". An operating system like Linux or Windows and their user-interface -- those aren't specific to any particular domain or particular use of computers. They can be used for pretty much any sort of use.

Algorithms and maths are "content free", in the sense that they aren't tied to any specific domains. Of course they need to be applied to the appropriate kinds of data, but usually this will not tie them to any specific domain area.

Pretty much all I've just said about technology applies to science and research as well.

A couple of years ago, I wrote a post The value of basic research that argues that the value of basic research is usually considered by looking at specific cases of basic research in isolation ("what problems does this help solve?", and "this research is done to satisfy curiosity"), but should instead be considered in broader terms (like how we can't predict the value of individual stocks, but can trust the stock market as a whole to increase in value, so it is with research. And the value of basic research is less in what problems it solves, but in the longer term of what it helps enable). But note that when I wrote that post, I didn't have this specific/broader angle clear in my head, so that is not explicitly mentioned in the post.

Another case where we need to consider broader details to properly understand technological development is described in this post: The value of greater convenience at scale in technological developments. It argues that the value of technology "merely" adding greater convenience is greatly underestimated when cases of that technology are considered in isolation, and that they must be seen in broader terms to appreciate their true value. (Related to this is what I wrote in The intuition that vertical technologies are the way to meet goals in computing).

Government policies, like economic or social policies, need to be understood in broader terms, too.

Too often people only consider these in terms of specific cases, considered in isolation.

Minimum wage policy, for example. People consider minimum wages in terms of their impact on individual workers. They don't look beyond that to broader details of the impact of the minimum wage level. They don't consider the impact on companies who will be employing a number of people at whatever the minimum wages are, over an extended period of time. How will that affect the company's hiring? What sort of impact will different policies have on the economy over larger timescales?

Probabilities (statistics) also need to be thought of in broader terms. Say there's a million-to-one odds of something happening to an individual. Perhaps the odds are of them having a certain medical condition. When we consider such odds and focus on a specific case like this person, it seems almost impossible that it could apply to them. I believe that most people tend to treat such low odds as equivalent to "impossible", for this reason. But this is the wrong way to look at such odds. We need to look at this issue in broader terms, and realise that probabilities apply to a "population". There may be one-in-a-million odds that a person has this condition, but there are almost 8 billion people on Earth, so there would likely be around 8000 people with the condition. Could this individual we're considering be one of those 8000? Of course that's possible, so we can't rule it out. 

Another example involving probabilities. Say that, again, there's a million-to-one odds of something happening to an individual. Like having a dream about an old acquaintance they haven't seen in decades and then the very next day running into them on the street. Lets imagine that the odds of this are a million-to-one. Many people, if such a thing happened to them, would consider it an amazing occurrence. It would almost seem like magic or evidence of the divine. But it only seems that way if you're considering the odds in terms of this specific case of this happening to you. If you take a broader 'population-level' perspective, there's nothing remarkable about it at all. Say there's 100 000 people (out of everyone on earth) each night have dreams about acquaintances they haven't seen in decades. And say that those people each walk past 100 other people each day. And that all this applies every day of every year. Given those numbers we'd expect 10 people, every single day, to have one of these one-in-a-million occurrences.  Of course the numbers I've used are all completely contrived, but I hope the point is clear.

Evolution is another phenomena that is seen in terms of a specific case considered in isolation, rather than in broader terms. When a lot of people think of evolution, they think of the concept of evolution. This is the 'specific case' when it comes to evolution. And they think of how that concept is defined (in terms of variation, heritability and selection). But the picture of evolution in terms of this definition is very different to the picture of it in broader terms, in terms of the kinds of concrete processes that occur in practice, that can span very long periods of time. To appreciate the broader view of it, we need to appreciate the huge scale it can operate on, the gradual nature of the changes, and things like the specific sorts of structure that can apply to the evolutionary processes, like evolutionarily stable strategies, kin selection, and sexual selection. It is only by being able to appreciate it in these broader terms that we can have good intuitions about evolution. I wrote a post about this here (though didn't put the matter in terms of specific-case vs broader details).

How can people get better at looking beyond specific cases? I think probably by learning about systemic things. By learning about evolution and economics, for example. I think learning mathematics and programming are useful too. They force you to think beyond specific cases to the general case. And there's simply making a conscious effort to take a broader perspective in all situations.

I think this ability should be taught in the education system. It's very important. It has such broad importance for thinking about so many different important topics. Rather than just teaching topics that help you think in this way, like evolution and economics, it should teach the specific principle of always considering the broader details, regardless of what the topic is.

.

Related: Chris Anderson's post The Probabilistic Age.