Emended definition—Muroid rodents in which the jugal is large, spanning most of the middle zygomatic arch (Major, 1897); the tongue retains three circumvallate papillae (Tullberg, 1899; Vorontsov, 1967); and the enamel face of the lower incisors bears two low and inconspicuous ridges, close-set in parallel and positioned mediolaterally (confirmed by MDC in all nine genera).

Group exceedingly diverse morphologically (see Carleton and Musser, 1984:341-2), but the above characters, in combination, uniquely define nesomyines among assemblages of living muroids. Jugal size and number of circumvallate papillae are probably primitive conditions, but the acquisition of various lower incisor ornamentations is thought derived (Flynn et al., 1985; L. D. Martin, 1980). Similar longitudinal incisor ridges are also found in certain cricetomyines among extant groups, and in the extinct Miocene African genera Afrocricetodon, Notocricetodon, and Protarsomys (Flynn et al., 1985; Lavocat, 1973). The morphological diversity of Madagascar’s native rodents has questioned their monophyletic origin and prompted numerous classificatory arrangements (see discussions in Carleton and Musser, 1984; Jansa and Carleton, 2003a). Proponents of a single ancestral origin have usually arranged nesomyines as a subfamily or tribe within Cricetidae (e.g., Miller and Gidley, 1918; Simpson, 1945; Vorontsov, 1959) or as a subfamily within a broadly defined Nesomyidae, which includes African groups such as cricetomyines, tachyoryctines, and Mystromys (Chaline et al., 1977; Lavocat, 1978). Ellerman (1941, 1949) was convinced that nesomyines are polyphyletic and dispersed the genera among five subfamilies of Muridae sensu lato.

Recent studies of DNA-DNA hybridization and genetic sequence data, although not finally resolving all aspects of the debate, persuasively demonstrate that Ellerman’s radical classificatory treatment is untenable. Surveys of few nesomyine taxa, using DNA hybridization or nuclear genes, have supported monophyly of Nesomyinae (DuBois et al., 1996; Michaux and Catzeflis, 2000; Michaux et al., 2001b); whereas, results based on dense taxonomic sampling, using a mitochondrial gene, have disclosed paraphyly, although most nesomyine genera were associated closely within two clades (Jansa et al., 1999). According to phylogenetic interpretations of nuclear genes, Michaux et al. (2001b) depicted Nesomyinae (two genera sampled) as a basal clade with other African groups (Cricetomyinae, Dendromurinae, Mystromyinae), a result that supplies some empirical support for Lavocat’s (1973, 1978) concept of a family Nesomyidae that embraces the archaic remnants of an African radiation dating to the early Miocene or late Oligocene (also see Carleton and Musser, 1984:344).

Living members endemic to Madagascar, known only from late Quaternary and subfossil Holocene deposits (McKenna and Bell, 1997). One Miocene occurrence, described from Africa, however, appreciably extends the fossil history of the group, if the identification proves true. Lavocat (1978) viewed Protarsomys, early Miocene of Kenya, as close to the ancestry of Malagasy Nesomyinae, and Chaline et al. (1977) emphasized that relationship by placing the Miocene fossil in synonymy under extant Macrotarsomys, as did F. Petter (1990). Others have questioned so close an affinity and specifically disputed their generic equivalence (Carleton and Goodman, 1996:246-249, 252). Macrotarsomys represents a morphology that evolved in situ, coincident with formation of the island’s dry western landscapes. Whether Protarsomys is a distant phylogenetic relative to extant nesomyines must await additional critical review of Miocene taxa and-or future discoveries. Brachyuromys, otherwise known only from the Recent of Madagascar, was mistakenly recorded from the late Miocene of Namibia (Conroy et al., 1992), together with Myocricetodon, Notocricetodon, and Protarsomys, but this fossil was subsequently reidentified as an extinct spalacid genus, Harasibomys (Mein et al., 2000a). Thomas (1915) in erecting the replacement name Majoria (for Myoryctes Major, 1908) uncritically stated that it represented "a fossil Madagascan rodent," an association just as uncritically perpetuated by others (Simpson, 1945; Vorontsov, 1959; Carleton and Musser, 1984); the holotype is an innominate bone of Plesiorycteropus, a member of the extinct order Bibymalagasia (see MacPhee, 1994).

Ellerman (1949) provided the first critical synopsis of nesomyine taxa, with subsequent descriptions and taxonomic arrangements updated by F. Petter (1972c, 1975a). Resurgence in field and museum studies over the past decade has vastly improved taxonomic, distributional, and ecological knowledge of nesomyines (Carleton, 1994; Carleton and Goodman, 1996, 1998, 2000; Carleton and Schmidt, 1990; Goodman and Carleton, 1996, 1998; Goodman and Soarimalala, 2002; Goodman et al., 1996, 1999b; Ryan et al., 1993; Soarimalala et al., 2001), as summarized for all species in Goodman and Benstead (2003). Conservation concerns associated with introduced Rattus populations and their impact upon native rodents discussed by Goodman (1995).