On the Indo-European reflexive

• Published in: Journal of Indo-European studies Vol. 26; no. 1-2; pp. 121 - 129
• Main Author: SHIELDS, K
• Format: Journal Article
• Language: English
• Published: Washington, DC Journal of Indo-European Studies 01.04.1998; Institute for the Study of Man
• Subjects: Ancient languages; Demonstratives; Diachrony and historical linguistics; Etymology; Indo European Languages; Language; Linguistics; Morphology; Reflexivity; Syntax; Radical; Etymology; Morphosyntax
• ISSN: 0092-2323; 2831-9060

[...]Cardona (1987), utilizing a morphological segmentation first proposed by Cowgill (1965: 169-170), considers *s(we)- as "consist[ing] of the third person pronominal *s- followed by the oppositional affix *we-, *wo" (4).5 It is not my purpose to critique each of these theories; instead, I merely want to present a reasonable alternative to those etymologies heretofore proposed. [...]se/o- represents "the contamination of [the zero grade] of *(e/o)s and the demonstrative *e/0- or ... the thematization of *(e/o)s" (Shields 1992: 29).6 Such contamination of demonstratives is a common phenomenon which is motivated by the fact "that by use a demonstrative tends to become weaker and weaker in its deictic force, and is therefore continually reinforced by being compounded with itself or with other demonstratives or with adverbs" (Lane 1961: 469). [...]multiple functions for a single deictic form are common. [...]in regard to Greek particles, Denniston (1966: lvi) observes: "... few Greek particles possess one meaning and one alone. What I would propose is that just prior to the dissolution of the Indo-European speech community, the proto-language marked both the reflexive and logophoric functions (like those attested in Latin) by means of the originally demonstrative stem * (e/o)s- and its contaminated variant in *se/o-, both of which had fully assumed the role as markers of coreferentiality. Since the language had available in its morphological inventory an indirect speech marker in *-we/o- (which was subject to contamination with *r), it merely followed a common typological path to differentiate the true reflexive and logophoric functions by affixing the indirect speech morpheme to *s(e/o)-, i.e. *s(e)-we-, when logophoric reference was to be expressed.8 Despite the emerging formal contrast between the reflexive and logophoric functions, this explicit opposition apparently did not become highly productive and was subsequently lost in the dialects, with the two markers emerging as functionally equivalent elements.