AN IMPORTANT HYPOTHESIS OF GENERATIVE GRAMMATICAL THEORY IN RECENT YEARS HAS BEEN THAT GRAMMATICAL TRANSFORMATIONS ARE MEANING-PRESERVING. THIS THESIS IS AN ATTEMPT TO SHOW THAT IF CERTAIN OTHER COMMON THEORETICAL ASSUMPTIONS ARE HELD CONSTANT, THAT HYPOTHESIS IS UNTENABLE. RULES OF WORD-DERIVATION--SUFFIXATION, PREFIXATION, COMPOUNDING--FREQUENTLY YIELD RESULTS WHOSE MEANING IS NOT PREDICTABLE FROM THEIR COMPONENT PARTS. IT IS ARGUED THAT SOME OF THESE MUST BE IN THE TRANSFORMATIONAL COMPONENT OF THE GRAMMAR OF ENGLISH. SOME DERIVATIONAL PATTERNS OF ENGLISH ARE ANALYZED IN DETAIL. ARGUMENTS ARE GIVEN THAT THE DERIVATION OF ADJECTIVES IN "-ABLE" MUST BE TRANSFORMATIONAL. NOMINALIZATIONS IN "-ITY" ARE SHOWN TO BE SEMANTICALLY ERRATIC. IT IS THEN ARGUED THAT THE FORMATION OF THESE MUST BE SUBSEQUENT IN THE GRAMMAR TO THE FORMATION OF "-ABLE" DERIVATIVES. ON THE ASSUMPTION THAT THE COMPONENTS OF THE GRAMMAR ARE HOMOGENEOUS, IT FOLLOWS THAT "-ITY" FORMATION MUST BE IN THE TRANSFORMATIONAL COMPONENT. IF RULES OF DERIVATION ARE IN THE TRANSFORMATIONAL COMPONENT, THEY MUST PARTICIPATE IN ITS ORDERING. SOME FACTS OF DERIVATION ARE EXAMINED IN THE LIGHT OF THIS CONSEQUENCE. ORDERING OF DERIVATIONAL RULES IS SHOWN TO OFFER A SIMPLE EXPLANATION OF SOME OTHERWISE PUZZLING PHENOMENA. IT IS THEN DEMONSTRATED THAT DERIVATIONAL RULES CANNOT BE LINEARLY ORDERED. A HYPOTHESIS IS PROPOSED AS TO THEIR ORDERING. THIS REPORT IS BASED ON A THESIS SUBMITTED AUGUST 1967 IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY AT THE MASSACHUSETTS INSTITUTE OF TECHNOLOGY. (AUTHOR/AMM)