An often noted convention of Dungeons & Dragons is the very limited selection of weapons available to the magic-user as compared to the fighting-man. Even the cleric, with full access to all types of armour, is prohibited from the use of the vast majority of martial implements, including bows, spears, and swords. This is somewhat at odds with the Chainmail Fantasy Supplement where it is asserted that "wizards can handle magical weaponry" (p. 30), that "certain magic-users can wield magical weapons" (p. 38), and also that Moorcock’s Elric of Melniboné "combines the attributes of the Hero-type with wizardry, and wields a magic sword in the balance!" (p. 38). Given such precedents, and the prominent depiction of a sword wielding magic-user in the well disseminated Lord of the Rings, it is perhaps unsurprising to learn that many early groups ignored the admonition in Men & Magic that magic-users should be restricted to the use of daggers (p. 6). Indeed, perhaps in response to a perceived prevalence of the variant, Gygax specifically singled out the practice in his Sorcerer's Scroll column as a point of contrast with prospective Advanced Dungeons & Dragons campaigns (Dragon #26, Vol. II, No. 12, p. 30).
Such changes between Chainmail and Men & Magic foreshadowed the general strengthening of fighters relative to magicians referred to in the Player’s Handbook (p. 7), a trend that was continued in Unearthed Arcana. Nonetheless, despite the definite archetypal division that Gygax ostensibly sought to establish between sorcerer and swordsmen for men, the same did not apply to elves or half-elves. The multiclass terms "fighter/magician", "magician/thief" and "fighter/magician/thief" were coined specifically for them, and blurred the lines between the classes. The counterbalance, that such characters have a limitation on the experience level that they can reach, has proven less than satisfactory for some players. The relaxing of those limits for second edition only exacerbated the issue, and provided no alternatives. As I recall, our group responded by banning multiclass characters altogether, haughtily declaring them to be only suitable for "munchkin" players. Only single class characters were allowed at our table, and woe to he who dared to mention the dualclassing rules.
Naturally enough, we strained against the boundaries we had set for ourselves. For my part, I petitioned the game master running our Greyhawk game to allow my first level elf magician to gain proficiency in the long sword and long bow, a proposition endorsed by the rest of the group. I also gleefully rolled up an elf priest of Corellon Larethian for our Forgotten Realms game using Monstrous Mythology, with the same result. The latter character was horribly killed by revenants somewhere in the underdark, but the former progressed as high as level nine or ten before I retired him, and I presume still lives somewhere in the vicinity of the free city. When Player's Option: Skills & Powers was announced with the tag line "ever wanted your wizard to wield a sword?", I thought that we had pre-empted the designers and natural progression of the game. I did not consider the possibility that this was a belated response to perceived changes in the expectations of their audience, and did not realise that it signalled a paradigm shift in the design of the brand. D20/3e was at the door.
When I resurrected my old second edition Advanced Dungeons & Dragons campaign to start a university group, one of the first things that had to be decided was exactly what rules would be used. The multiclass and dualclass rules remained unattractive to me, but I still wanted to allow for some flexibility in character concepts. Using the various Player’s Supplements and Player’s Options books in combination with the Player’s Handbook and Dungeon Master’s Guide, I came up with some rudimentary rules for creating variant classes for my newly recruited players. We ended up with a human fighter (knight), a half-orc fighter (barbarian), a dwarf cleric (axes allowed), and a human magician/thief (spellthief); they were joined a few sessions later by an elf fighter/thief (scout). It should perhaps have been apparent to me then that I had wasted rather a lot of time, but in truth it was a learning process. Players naturally gravitate towards the basic classes for a variety of reasons; preconceived expectations and the palatability of familiar conventions no doubt play their part, but also because the archetypes are so general that they are difficult, if not impossible, to escape. Indeed, classes are perhaps best understood as tools intended to facilitate play; what differentiates one fighter from another are the details of his life, the adventures he has had, and the personality he projects, not signature weapons, nor detailed rules governing and defining his every ability.
But I digress. One of the more interesting facets that I encountered in the process of creating variant classes is the distinct lack of appropriate terms for combination classes. Typically, new subclasses are a more specific subset of the class from which they derive: pirate, thief, borderer, reaver, ranger, knight, barbarian, hoplite, enchanter, necromancer, warlock, acolyte, ascetic, sorcerer, invoker, etcetera. Suitable designations for describing the fighter/magician and magician/thief are few and far between, never mind their subsets. The D20/3e designers clearly encountered this problem when adding new classes, and made use of compound nouns patterned on the model of the "spellthief" and spellblade", resulting in nomenclature such as "warmage", "duskblade" and "hexblade". More natural sounding designations were less frequently conceived, such as the "beguiler" (a sort of illusionist/thief). The likely reason that we lack an abundance of adjectives and nouns to simultaneously imply both magic and fighting, or magic and theft, is that the assumption of their individual exclusivity is a conceit inferred from the concept of discrete classes. That is to say, Elric of Melniboné is a swordsman and a sorcerer; a spellsword is a game construct.
Which brings us to a more pertinent question: by exactly what means does a magician gain his powers? In mythology and fiction, the remarkability of a hero is often prefigured by a strange conception or ancestry. Achilles, Hercules, Alexander, and Arthur, as well as Bilbo, Aragorn and many others can be counted amongst these. The magical powers of Merlin are often ascribed to a demonic parentage, as are those of the sorcerer Tsotha in The Scarlet Citadel. It is similarly notable that in Advanced Dungeons & Dragons the races eligible for the fighter/magician and magician/thief combinations are elves and half-elves, or fey and fey touched. Whilst the game certainly posits that magicians learn their sorceries, and that natural ability in the form of intelligence governs their potential, this does not preclude the idea that some other condition, perhaps natural talent or strange heritage, is required for the study of magic to be successfully undertaken. Indeed, it might even be argued that to some degree this is assumed to be the case, else the world would surely be filled with sorcerers. Regardless, in a fantasy milieu where wizardry is taught and learned there must be those who, like the Grey Mouser, are displaced or otherwise choose not to devote themselves entirely to the pursuit of magic. It would perhaps be simplest to allow humans to multiclass as half-elves, but characters in my campaign rarely advance far beyond ninth level, and truth to tell I am still ill disposed towards the multiclass rules. So, on the supposition that others might feel similarly, here are two subclasses I use written up for OSRIC in pdf form.