Scrivener of Doom
Adventurer
(snip) The first edition monster manual introduced us to 'Type 1' through 'Type VIII' demons. There are several aspects of this presentation where I think the ultimate intention was lost, and where I also think Gygax thought better of his original design. The first thing that I think was lost, was Gygax never intended the demons to be hierarchical, with Type 1 being the weakest and ascending to stronger and stronger types. A careful reading of the 1e AD&D stats does not show a linear progression in power from Type 1 to Type 8. There is an overall progression, but some lower numbered types have more HD or other superior abilities to higher numbered types. The further we've gotten from the original MM, the more hierarchy has been introduced. And the second thing that I think was lost was that I don't think Gygax really intended any particular type to be numerous and pervasive. Each type was I think intended to represent a relatively small number of individuals, many of which would be named. I think he intended there to be an infinite number of types, which by now might be publishing type 60's or type 90's with the clear indication that there were many more types not yet described.
By the time the DMG was written, I think Gygax had hit upon a better scheme than publishing stat blocks for demons, and that appeared along with several other ideas that he'd hit upon between writing the MM and the DMG, in the Appendixes. In the Appendix, he presents a template for generating a very large number of random fiendish beasties, and I think that and not what we ended up with, is what should have been the model for future demons. (snip)
1. The 1E Monster Manual included type I to type VI demons; there were no type VII or type VIII demons (except perhaps in Arduin Grimoire?).
2. Type I-VI demons were actually introduced in OD&D's Eldritch Wizardry.
3. On the contrary, it is clear that Gygax (well, as clear as anything could be with Gygax who loved the non sequitur school of game "design") did mean for there to be a demonic hierarchy as evidenced by the rankings of same from type I to type VI. Further, the hit dice issue is not as clear cut as the 1E Monster Manual would have you believe because, if you go back to the Eldritch Wizardry source, Gygax intended for the type IV and type VI demons to have lower hit dice but those hit dice being a higher value. (Demogorgon and Orcus were also noted as using the higher hit die value.) Why did he do this? Who really knows. I just assume it was yet another non sequitur that he never had time to develop or explain and that his editor was told to ignore.
4. I seem to recall that the random tables in the DMG were first published in The Dragon (as it was). I've gone through Dragondex and some of PDFs but cannot find if my memory is faulty or not. It would be interesting to read the commentary that goes with the introduction of these tables to see if there was any underlying logic or if it was just something interesting to publish. I tend to agree that a creature type that is supposed to be the exemplar of chaotic evil should be a lot more chaotic than its carefully structured hierarchy of power would otherwise lead us to believe. Monster Manual II's hordelings/hordlings were actually a much better representation of chaotic evil paragons than Gygax's demons (despite being, IIRC, neutral evil).