In mathematics and theoretical computer science, theoreticians seem to always be grasping for synonyms of the word “type.” This post serves as a convenient resource for these words.
Types of “Type”s
| analogy | types of types |
|---|---|
| abstract | type (type theory), kind (type theory), set (set theory), collection, variety, category (category theory), group (abstract algebra), lot, genre, division, class (OOP), circle, assortment, aggregation, batch, bundle, bunch, conglomerate, cluster (statistics), sample (statistics), gathering, level (type theory), share (economics), chunk (software), fragment (software), partition (software, set theory), grouping, configuration, conformation, design, framework, methodology, derivation, selection, classification, frame, pattern (type theory), theory (category theory), topic, span, sphere, affinity, stretch, case (type theory) , homology (category theory), sort (type theory) |
| social/political | cast, gang, band, club, faction, party, assembly, cartel, clique, covery, covey, league, suite, rank, tier, sector, department, standard, echelon, school, arragenement, organization, system, constitution, alignment, formation, cult, structure, religion, history, field, domain, estate, house, dynasty, empire, rule, reign, government, administration, jurisdiction, management, legion, dominion, regimen, tenure, order, kingdom, varna, association, alliance |
| artistic | style, orchestration/orchestra, ensemble, band, sinfonietta, choir, chorus, chorale, verse, hymn, paean, song, production, composition, tune, theme, motif, subject |
| biological/genetic | strain, ilk, species, breed, family, ancestry, generation, kin, progeny, legacy, lineage, brood, kindred, folk, genealogy, pedigree, stock, tribe, descent, scions, issue, organism, parentage, stemma, genus, phylum, kith, cognate |
| natural/temporal | bag, flock, herd, pack, pool, branch, world, tree, realm, dimension, place, age, epoch, era, period, sheave (category theory), fiber (abstract algebra), glob (category theory), space, site (category theory) |
| technical | grade, caliber, gradation |
Criteria
In choosing these words, I mostly looked to mathematical convention for what works as new terms. Though my goal is that these words should be able to be used as names for new mathematical/computical structures, I do not suggest that each word is the most proper fit for any new structure. The same is true in convention, even with the most general terms. For example, “group” could probably not have been suggested in place of “type,” because “group” insinuates a sort of relatability between the elements of the group, where terms of types are not necessarily related in any way. The term “group” fits mathematical groups because the elements of a mathematical group do indeed relate in that they can be composed to get other elements of the group and so on. In the same way, the word “system” could be used well for some mathematica/computical structures and not others.
Of course, on the other hand, there are interchangable terms. For example, “type” and “kind” could probably have been interchanged without much trouble, renaming “type theory” to “kind theory,” which is perhaps a more friendly and less divisive name even.
So, the criteria below are meant to be very broad in order to capture any words that have any non-negligible likelihood (in my opinion) of dubbing a mathematical/computical structure.
- the word refers to several things
- the word need not be very specific about the absolute qualities of the referred things
- the word dilineates the things from other things in a simple, context-free way
- there is potential for peculiar connotations of the word to translate into some metaphorical metaphor describing the strucutre
The last point is important for a large class of my suggestions, because plenty of the words I admit have connotations that might seem to contradict the second point about not being specific about absolute qualities. For example, how is “genealogy” not specific about genetic qualities? The words I suggest are chosen such that even if they do have more specific connotations, they are easily generalizable. With “genealogy,” it is intuitive to generalize the idea of the original genetic “genealogy” to the genealogy of inheritance in OOP or the genealogy of increasingly-axiomatized structures in set theory, etc. To take this example one step further, here’s a suggestion for what a mathematical genealogy could be defined as:
A genealogy is a directed graph of theories, where there is an edge from theory \(A\) to theory \(B\) if the axioms of theory \(A\) are a subset of the axioms of theory \(B\).
I haven’t noticed a very common definition of “theory” in mathematical texts, though I’m sure its out there somewhere. This is how I define it:
A theory is a set of propositions. A set of propositions are considered axioms of a theory if all the propositions in the theory can be reached by a finite number of applications of the axioms, via the rules of logical inference.
Case Studies
Type Theory
As a particular example close to my interests, the field of theoretical computer science type theory revolves around a concept called “type,” where data manipulated by a computer can be annotated with data types such as “integer,” “boolean,” “string,” etc. In this sense, a datum’s type corresponds to a way of reading the raw data; the raw data can all be considered to be of a single trivial type (e.g. “bitstring” or even just “computer memory”).
Of couse, these are particularly basic examples of data types. Type theory is mainly interested in formalizing systems of interesting types, such as polymorphic types, inductive data types, dependent types, or even self types. However, to investigate such interesting type systems, a higher-order concept of type is needed. For example, what is the type of a type, if the type is considered as a type of data type itself? A common name for the type of a type is “kind” i.e. types are to data (or, terms) as kinds are to types. Then we may go further – what is the type of a kind? There is a common name here too: “sort”. Altogether:
- type is to term as kind is to type i.e. the type of a type is a kind
- sort is to kind as kind is to type i.e. the type of a kind is a sort
What lies beyond? Typically, as far as I can tell, “sort” applies well to the infinite heirarchy of types of types of … of sorts. A little underwhelming. At this point many systems (such as Coq and Agda) use a numbering scheme to distinguish levels of the heirarchy. With these retroactive renamings:
- type := type0
- kind := type1
- sort := type2
- type of type of … := typeN
Abstract Algebra
Abstract algebra upon set theory with a (non-technical) variety of structures, the most commonly-recognized name among them being “group.” The name is fitting because it is so generic a name to give to such a specific mathematical structure, yet one that is a very generalizable itself. Notably, “an introduction to abstract algebra” is often shortened to just “group theory.” There are other structures beyond groups, of course. And though their names seem just an unspecific, it is fatal to mix them up. A summary of abstract algebra is: the domain of interesting structures generated within the axioms of set theory. See [algebraic structures] for an extensive list. Here is an interesting selection:
- pointed set
- magma
- semigroup
- monoid
- group
- ring
- lattice
- field
- algebra
- arithmetic
- module
- vector space
- topology
- space
Category Theory
When modern mathematics was first formalized in the 19th and 20th centuries, a new standard basis for all of mathematics was formalized: set theory. The basis of set theory, of course, was the set. While in layman’s terms a set of things might be just another way to say a collection of things or a group of things, set theory gave a technical definition to “set” to be used in this formal context. Given as a basis for all of mathematics, it was plausible at the time that any mathematical theory could be expressed (and many believed that, beyond just expressing any theory, any theory could also be proven or disproven) in the language of sets. Famously, Alfred Whitehead and Bertrand Russel attempted to prove this plausibility in the Principia Mathematica.
Since those developments however, mathematicians learned that there are some mathematical theories that are “beyond set theory” i.e. mathematical theories which represent structures that cannot be formalized in set theory. The prime example is category theory – the study of categories which turn out to be strictly more general than mere sets. Why the name “category”?; is it not just another term for “set”? Yes, it does seem to be. Additionally, it turns out that there is a correspondence between categories (from category theory) and types (from type theory), which is very convenient since “category” and “type” are roughly synonyms in the first place!
And once you you have new structure names like “category” floating around, of course many others come:
Of course, coming up with many different names that often are different words for “type” is not unique to category theory. But given that category theory usually is the most abstracty of the nonsense, it really has to start reaching for new words to name structures that are so general their intuitive structure is hard to grasp in a term other than “type” in a slightly specific context.