The author considers anew the origin of the Slavic suffix *-tajь, taking into account new Tocharian data which feature the agentive suffix -tau (e.g. Toch. B. olyitau ‘boatman’ : Toch. AB olyi; Toch. B käryorttau ‘trader, merchant’ : karyor ‘buying, business negotiation’), as well as the iterative-frequentative feature of the verbal suffix *-teh₂- in the Indo-European languages. The iterative-frequentative aspect of the Indo-European suffix *-teh₂- is securely preserved in the Latin verbal system, cf. Lat. eō, īre ‘to go, walk, move, pass’, Gk. εἶμι ‘id.’ (< PIE. *h₁ei- ‘to go’) vs. Lat. itō, itāre (verbum iterativum vel intensivum) ‘to go, march’, Gk. ἰτητέον (adiectivum verbale) (< PIE. *h₁i-teh₂- ‘to go frequently’). It is suggested that the iterative-frequentative (and perhaps intensive) meaning of the suffix *-teh₂- was adopted from Indo-European verbal formations and introduced into a number of nominal forms, e.g. agent nouns (nomina agentis) with the (verbal) suffix *-teh₂-, e.g. PIE. *h₂erh₃-i̯e-ti ‘he ploughs’ PIE. *h₂erh₃-teh₂-i̯e-ti ‘he frequently (or constantly) ploughs’ PIE. *h₂erh₃-teh₂-s m. ‘a man who frequently (or constantly) ploughs the earth’, i.e. ‘ploughman, farmer’ PIE. dial. *h₂erh₃-teh₂-i̯o-s m. ‘id.’. The author concludes that the Proto-Indo-European archetype *h₂erh₃-teh₂-(i̯o)-s originally denoted ‘a person who frequently (or constantly) ploughs the earth’. Put differently, the Indo-European nominal suffix *-teh₂-, attested in certain agent nouns in Baltic, Greek, Slavic and Tocharian, was characterized by the iterative-frequentative aspect taken over from the corresponding verbs in *-teh₂-. The original semantic difference, reconstructible for the Indo-European proto-language, has been completely forgotten in most of the daughter languages. This is why the Ancient Greek noun ἀρότης m. ‘plougman, farmer’ (< PIE. *h₂erh₃-téh₂-s m.), which originally denoted ‘a person who frequently or constantly ploughs the earth’, seems to be fully synonymous with ἀροτήρ m. ‘plougman, farmer’ (< PIE. *h₂erh₃-tér-s m.) which originally indicated a man who is ploughing currently but not constantly. The same semantic difference must have existed in Baltic (e.g. Lith. artójas ‘ploughman, farmer’, OPruss. artoys ‘farmer’ vs. Lith. arėjas m. ‘plougman’, Latv. arẽjs m. ‘ploughman, farmer’), as well as in Slavic (e.g. Pol. rataj ‘ploughman, farmer’ vs. oracz m. ‘ploughman’).