In set theory, a tree is usually defined as a partial order such that the set of elements below any given one is well-ordered. I am interested in the class of partial orders $$P$$ such that for every $$p \in P$$, the set of $$q \leq p$$ is just linearly ordered. Does this have a name?

• rd.springer.com/article/10.1007/BF00571186 – Asaf Karagila Jan 18 at 15:45
• Prefix orders seem to be relevant, if for no other reason than they appear to fit the required definitional niche. See en.m.wikipedia.org/wiki/Prefix_order – Not Mike Jan 18 at 17:18
• Thanks @NotMike! It's nice to see this coming from outside of pure set theory. Feel free to put this as an answer. – Monroe Eskew Jan 18 at 17:26
• Adeleke and Neumann have a Memoir of the AMS 'Relations related to betweenness' which considers various structures along these lines. In particular, a partially ordered set satisfying the property above which is also assumed to be connected (every pair of elements has a common lower bound) is called a semilinearly ordered set. I don't know if they have a name for such objects if they're not connected. – shane.orourke Jan 18 at 19:22

They are also called trees.

In that terminology, trees of your first kind are known as the well-founded trees, since they are trees where the tree order is well-founded (and well-founded linear orders are the same as well-orders).

I think that the situation is that because set theorists are mainly interested in the well-founded case, the terminology evolved to drop the adjective from well-founded trees.

There are many competing definitions of tree in mathematics, not all equivalent. For graph-theorists, for example, a tree is a certain kind of cycle-free graph.

• This is also what I thought off the top of my head, but Jech, Kunen, and Kanamori all put well-foundedness into the definition of a tree. Do you know of a reference the defines trees as the more general class? – Monroe Eskew Jan 18 at 15:32
• I guess the complication also is that "well-founded tree" now means something else in set theory, where the tree is growing downward and has no infinite branches. – Joel David Hamkins Jan 18 at 17:58
• For one reference: Keisler uses the linear order definition in his article "Models with tree structures". Full citation: H. Jerome Keisler. “Models with tree structures”. In: Proceedings of the Tarski Symposium (Proc. Sympos. Pure Math., Vol. XXV, Univ. California, Berkeley, Calif., 1971) . Amer. Math. Soc., Providence, R.I., 1974, pp. 331–348. – Kameryn Williams Jan 18 at 22:27
• But set-theoreticians, who use the word "tree" to mean well-founded tree, must occasionally have to refer to the more general trees. What do they call them? If I use "tree" to mean a poset in which the predecessors of any element are well-ordered, then what should I call a poset in which the predecessors of any element are linearly ordered? I have some vague recollection of hearing them called "pseudotrees" or something like that. – bof Jan 23 at 5:58
• @bof A colleague pointed me to several set-theory papers by Koppelberg-Monk, Bekkali, and Alos-Ferrer and Ritzberger, where the term pseudotree is used for exactly this kind of poset. – Monroe Eskew Jan 24 at 9:36

After a little bit of searching, the notion of prefix order seems to be relevant; if for no other reason than that it appears to fit the required definitional niche.

(Also, it seemed worth pointing out the notion of prefix-order is precisely that of a "first-order tree".)