Setting tableaux / downward-branching trees
The
task is to set proof-trees in the style of the illustrated tree (which proves
the invalidity of the inference
--
cf. e.g. Jeffrey's Formal Logic or my own Introduction to Formal
Logic). [Oops, excuse missing negation sign!]
The illustration was set using the pst-tree package which is part of the pstricks family of packages. This was recommended as the best option in a recent exchange on comp.text.tex (see below).
This is a little cumbersome to use but produces quite nice results -- and it shouldn't be too difficult to define commands to wrap up some of the trickery. (However, at the moment, I couldn't easily get the sort of control over the angle of branching that I'd ideally like.)
The following package written for typesetting linguists' syntactic trees can be adapted to logicians' purposes, though my attempts weren't terrifically successful.
However, I've now found two useful pages
which show what can be done with qtree.
There is also an "unreleased" package which tweaks proof.sty (described here) to turn things upside down, which therefore produces a tableau proof with inferences marked by horizontal lines rather than sloping lines.
As noted above, the question of how to set Jeffrey-style treeswas recently asked on the usenet newsgroup comp.text.tex (22 June 2004). Replies include:
You could try pst-node or pst-tree
Herbert Voss (Herbert.Voss@gmx.net) |
| I suppose that I would use the pict2e package for that job Jose Carlos Santos (jcsantos@fc.up.pt) |
| You have some options:
(guido@guido.au)
|
The initiator of the thread replied
| Indeed, pst-tree works nicely. |
One day, I'll need to sort this one out for my own purposes fairly soon (I used FrameMaker for my own logic book, if you are wondering how that was set). When I do, I'll post again here.