Comment by tomrod | Hacker News Reader

Comment by tomrod | Hacker News Reader

Parent Full thread

Ultimately, many people don't understand what a function, mapping, transmutation, transcription, reclassification, or other process which takes items from one category and puts them in a new set of categories.

Programmers, especially in the data space, perform this activity _all the time_. It's almost as deeply ingrained as breathing. Many people who work in tech-adjacent areas get some kind of understanding of this as "business rules." But on all these fronts it is the application of mathematics, the logic and language of the structure of things.

For me, mathematics was always something that I was decent-but-limited at until I got a great teacher -- and then it all just clicked. I had this teacher at the early undergraduate level, but it was because of their absolute practicality that it all made sense. Prior instructors had focused on the _how_ to perform activities, but never really dug into the _what_ / definition of the activity. What the goal of the action was, and more importantly, how it could then be used as a tool in the toolkit.

What I have realized over time is that when things are difficult to explain, I find success by channeling the great math instructors I had in life.

I feel strongly that this relates to the DNA→RNA issue identified here. Simple abstraction, a game, you name it can help engage the _concept_ of reclassification, and then it becomes much simpler to apply the abstract concept to the specific process.

I've never heard of transmutation, transcription, reclassification, or category as data processing jargon.

Transmutation is for nuclear physicists, transcription is for court reporters, reclassification is for national security, and everything is categorical because people like the word.

How would they help with biology?

I have had some experience (among non-technical, non-bio co-workers) with what I suspect is the basic issue here (and I think the document link confirms this) - people just equate an "explanation" to connecting things with arrows and everybody kind of nods along while never agreeing on what the arrows mean.

It can be seen as antisocial if you - the recipient of the "knowledge" - question what they mean.

> [transmutation] ... as data processing jargon

Nor have I for that specific context. As Whitman noted, people are large and contain multitudes. Reclassification and transcription are certainly common enough colloquialism even if not formal processes. It's all mappings[0] in the end.

> everybody kind of nods along while never agreeing on what the arrows mean.

Yes, this is what I said, and elucidated an abstraction process to convey complex technical information I both experienced and have used in practice.

[0] https://en.wikipedia.org/wiki/Map_(mathematics)

I must be missing some important part of what you were trying to communicate.

It sounds as though you are saying "colloquialism[s]" have some value or relevance to technical understanding of a process.

It seems to me this is the primary means by which information gets lost because people think they know what is meant, but they don't.

It's the stumbling block that always reoccurs.

---- If someone says "map" and means function, then I know what a function is (I think).

If someone says "map" and means a higher order function, like in programming, then that's also something I think I know.

If someone wants to talk about morphisms that are not functions, then that sounds like it may be beyond my ability to follow.

But in any of the three cases, using the word "map" out of context fails to tell me which it is. Other associations, like the paper rand-mcnally map that I once used to drive across the country, are not helpful.

The point is that the physical process is not like programming or math -- there is a difference between "A becomes B" and "A is used to as a reference when creating B", which is conflated all the time in programming unless you are using linear types, which is rare. (Mutating a data structure doesn't give you a new type. Garbage collection does, but that's not part of the semantic model of a programming language).

Well, transcription is a the transference of information between mediums. Spoken to written. Spoken to recorded media. Written to other written. But strangley not written to spoken, so the analogy breaks down there.

But if you have the understanding that transcription is a copying to another media, then the jargon makes perfect sense. People used to dictate and have a secretary transcribe. But we don't do that any more. And like the sound of a typewrighter the verb "transcribe" has drifted out of common parlance and might therefor gone a bit stale as an analogy.

> everything is categorical because people like the word

> people just equate an "explanation" to connecting things with arrows and everybody kind of nods along while never agreeing on what the arrows mean

I bet you'd love category theory!

I can state categorically that I have read a little bit on category theory and neither fell in love nor remember it.

”Transform” is commonly used in the context of data wrangling.

Which is exactly the same wrong metaphor that the article is talking about. A transformation in data processing is a transcription in biochemistry creating something new without destroying the original), not a transformation in biochemistry (which destroys the original).

Reply on news.ycombinator.com