Hacker News Reader

Comment by dopylitty | Hacker News Reader

Whenever I see this need for somewhat arbitrarily classifying things into categories I wonder what the purpose is.

For example there was a long period of time when viruses were classified by the diseases they caused. So you had enteroviruses that cause GI diseases and rhinoviruses that cause respiratory diseases. When gene sequencing became possible it turned out a lot of these entero and rhino viruses are closely related, something that has implications for how they can be treated and how cross-immunity works. Meanwhile the various hepatitis viruses aren’t as closely related. So for viruses a useful classification might be based on which treatments work or which viruses elicit cross immunity.

So getting back to classifying dinosaurs and dwarf planets the question is what value the classification brings. Does classifying them make them easier to study somehow?

View on HN

>arbitrarily classifying things into categories

Most things have continuous features. Any kind of classification is of course arbitrary then.

It's arbitrary to distinguish these 3 types of T-Rex as well as it being arbitrary to draw a line between chicken and velociraptors.

Of course there's some point where most people agree that a classification makes sense, because there's a "well d'uh" kind of difference, like the chicken and the velociraptor. At the other extreme you'd classify every individual, because they're all unique. People will endlessly argue where to draw this line. There just is no answer to the question how far we should go.

Simpler classifications are useful to enable communication on a broad level. I'd be disappointed to go to a hip-hop concert even though I expected techno. If you told me some subgenre from hip-hop however I might not have understood what the music is they played, because it's not my thing.

Fine classifications are useful to enable communication between enthusiasts/experts. The hip-hop heads will have their subgenres, and the techno people, too. There are distinctions which don't matter to outsiders, but they do matter to insiders.

The finer the classifications are, the less useful their distinction, and the more likely that your communication partner will not understand you though. Techno now has so many subgenres, that I'd be hard-pressed to explain a lot of them.

The problem is that nobody can meaningfully draw a line.

You stated:

> Of course there's some point where most people agree that a classification makes sense, because there's a "well d'uh" kind of difference, like the chicken and the velociraptor. At the other extreme you'd classify every individual, because they're all unique. People will endlessly argue where to draw this line. There just is no answer to the question how far we should go. > > ... > > The problem is that nobody can meaningfully draw a line.

Your statement makes no sense. Anyone can draw a line between a chicken and a velociraptor. Anyone who argues that no line can be drawn between two etxremes are either deluded, insane, or with an agenda.

For purposes of scientific classification, the lines may blur sometimes. But the evidence at hand and discussions among experts does not seem to justify 3 different species. If so, one can only presume that those who wish there to be 3 species are finding ways to inflate/publish a paper and to claim a scientific stake, rather than there being some genuine need for classification.

>Your statement makes no sense. Anyone can draw a line between a chicken and a velociraptor.

Anyone can draw lines, that's not what I was saying tough, the chicken and velocitaptor example was exactly to show one of these very obvious distinctions.

The problem is rather where and how many lines to draw. No number of classifications is "correct" since a classification is only one of almost infinite possibilities to choose from.

If you treat things as forking branches in a tree, then refinement of branching in deep time, is informing of why modern developments diverge. You might have 4 things held to have a common ancestor and uncover 3 of then from one lineage and 1 from an earlier fork. This may for instance explain why Chimps are remarkably non-verbal for otherwise smart creatures.

It's why we're confident about the lineage to Homo Sapiens but the detail of Denisovan, Neandertal is huge. What if we unpacked a pre-lingual ancestor? What might it say if they were not toolmakers beyond the crow-trivial level, and what comes after were? Would we be able to tie tool making to vocalisation?

I don't think this is the only reason, and I have no belief T-Rex progeny do more than inform marginally KFC cooking times and relative fat content of duck and goose..

It's lumpers vs splitters again. https://en.wikipedia.org/wiki/Lumpers_and_splitters

> Whenever I see this need for somewhat arbitrarily classifying things into categories I wonder what the purpose is.

The purpose is to get more money. Imagine they have 3 T. Rex skeletons. Then instead of 3 T. Rex, they have a T. Rex, a T. King and a T. Czar much better than the shop around the corner.

In code terms, knowing the history of a codebase is often more important than knowing the current feature set it provides.

Reply on news.ycombinator.com