In this post, I revisit the general issue of analysis that I discussed in my previous post. There is a measure of overlap because I’m really searching for a way to communicate both the fundamental simplicity of analysis with all its potential complexity. Maybe the general principle for this post is that analysis is, at its roots, a simple intellectual action: dividing something into different parts, but that this simple action inevitably leads to increasing complexity.
As with so many things in which analysis is involved, this post started out simpler and shorter than it has become. My original plan was to write one short post that just did a better job of explaining the ideas in the previous post. But then, as I thought more closely about it, I found issues that hadn’t been discussed in my previous. It’s now looking like this will be a series of posts—at least two: this one will discuss the big idea of analysis and relatively simple, everyday examples; the next will look at some examples more closely, in hopes that they feel more like an academic example. I suspect that may end up as two or more posts. In a way, this story encapsulates an aspect of analysis in practice that I want to emphasize here: the more you do it, the more complexity you see, and that leads to expanding projects, that must be reined in for purely practical reasons: basically, if you want to finish a project, you have to stop analyzing everything. (And as I write that, I wonder whether I haven’t sparked the foundation for a third post: how do you stop analyzing once you’ve started. It’s an idea that I touch on briefly in the second post, but maybe it deserves its own? I’ll have to think about that…)
What is “Analysis”
At its root (its etymological foundations), “Analysis” is derived through medieval Latin from the Greek for “unloose” or “take apart.” (In contrast to “synthesis” whose roots lie in the Greek for “put together.”) This sense is generally in line with how the word might get used in a conversation. For example, after [a movie/a TV show/a meal at a restaurant], if one person is talking at length criticizing details of the [movie/etc.], the other might get exasperated and say “Stop picking it apart,” or “stop over-analyzing it.”
It is this basic “picking apart” that concerns me in these posts. It is a basic principle that can manifest informally (as a person might do with a movie/etc.) or one that can manifest as extremely detailed and formalized systems of analysis, as with psychoanalysis, or statistical analysis, or data analysis, or any other field that uses “analysis” in a title.
We Do It Automatically
The kind of analysis that is important in research (and other intellectual work) is something that humans do naturally and automatically—often without even noticing that we’re analyzing.
To apply it in research is to take an automatic, unconscious ability and work to make it conscious and explicit. Splitting things into pieces—into different parts or different aspects—is pretty easy. But making those divisions explicit is hard because of the complexity that tends to develop.
We all automatically split things up into different parts, which is reflected in our languages (including words like “parts,” “pieces,” “components,” “elements,” etc.) and much of our daily lives. We separate the world into all sorts of different categories. We eat food, which includes fruit, vegetables, meat, etc. We work, but have many different kinds of work: homework, housework, yard work, not to mention jobs, which are work. We separate the good from the bad. We divide people up into different groups: family, friends, acquaintances, people we don’t know, etc.
It’s true that many of these divisions are learned, but that doesn’t mean that we don’t naturally make divisions of some sort.
Analysis: Examples
Consider an apple. It is a whole in itself, but we pretty naturally separate it into a few different parts: stem, skin, flesh, core, seeds. Our basic sensory apparatus provides distinguishing information: stem, seed, and flesh taste different, smell different, look different, and feel different. Our basic sensory apparatus is already providing us information about differences in the world that lead to analysis of the apple into its different parts.
Consider a movie. It is a whole in itself, but we can easily divide it in many different ways that are familiar to cinephiles. We can say “The acting was pretty good, but the script was weak.” Or “The cinematography is great, the writing is great, the direction is ok, but the star annoys me, so I had trouble enjoying it.” We might like what we see (“great cinematography!”), but not what we hear (“poorly written dialogue”). We might like one actor and not another. Again, this is analysis in action, although few would think of this kind of thing as analysis. Unless we were to really get into a lengthy discussion of different aspects of a movie, and then someone might say “stop analyzing it! You’re ruining it for me!”
Research and Analysis
Research takes this basic ability to distinguish between things and tries to make it explicit and formal. For the researcher, it’s not enough to say that it’s obvious that you have stem, seeds, and flesh, or acting, directing, writing, and cinematography. It’s necessary to begin to formalize.
Formalized analysis is crucial in research because it allows a research community to work together. Researchers who doesn’t explicitly express their analyses can’t have their researcher reviewed or trusted by others. The need to share and provide explanations and evidence that can be examined leads to detailed discussions (articles books, etc.) that can themselves be analyzed (and will be by other researchers who will look for strengths on which to build and weaknesses to correct).
In practice, research communities develop different analytical frameworks and methods of analysis as a result of the attempt to explain and examine each others’ work. These become increasingly detailed and complex over time, as each successive generation of researchers turns their analytical abilities to the questions of interest. Sometimes entirely new analytical frameworks develop, but these, too, are subject to close examination that leads to complex formal analytical systems.
Psychoanalysis, for example, depends on familiar analytical divisions: the id, ego, and super-ego represent parts of a large whole. So, too, the conscious and unconscious. Each different pathology is a part of the large whole of “poor mental health.” And each pathology itself is distinguished by a number of different characteristics that are parts of the pathology. To become a psychoanalyst, is to adopt a specific set of analytical frameworks regarding the psychology of individuals and the nature of psychotherapy as well. Other theories of psychology and psychotherapy may not be called “psychoanalysis,” but they too adopt different analytical frameworks.
Mathematical analyses separate the world into different symbols that represent different parts of the world and distinct relationships between the parts. Physics, of course, presents the interactions of objects in the world as a set of symbols and mathematical equations. In a business setting, the large-scale system of a factory, for example, might get represented in mathematical equations that separate out machines that produce goods, goods that are produced, rates of production, costs of production, necessary workers, etc.
Conclusion
Analysis happens. If you examine something closely—an object, an interaction, an idea—you will begin to distinguish different aspects or parts of it. These distinctions are analysis. To move that analysis into an academic or research setting really only requires that you try to make your analyses explicit as you develop them, so that they can be examined for flaws (by you and by others).
Of course, making analyses explicit and then looking at those analyses with an eye for flaws may be a path to good research, but it is not a path to simplicity.
I’m going to close here and in my next post (or posts), I’ll look with greater detail at some examples to show different ways in which things can be analyzed and to discuss the expansion of complexity, which can be both good and bad.