This family contains species of fossorial and subterranean muroids arranged in the Myospalacinae (zokors, Eospalax and Myospalax), Rhizomyinae (bamboo rats, Cannomys and Rhizomys), Spalacinae (blind mole rats, Spalax), and Tachyoryctinae (African mole rats, Tachyoryctes). All extant species are characterized by extreme morphological, physiological, and behavioral specializations associated with subterranean life in tubular burrows (Gambaryan, 1960; Gambaryan and Gasc, 1993; Nevo et al., 2001; Tullberg, 1899). Although each subfamily can be readily diagnosed by unique traits (Carleton and Musser, 1984), and although molar occlusal patterns among living species are dissimilar, all share a comparable cranial and postcranial skeletal architecture integrated with a myological system that characterizes highly specialized fossorial rodents. Such a phenotypic resemblance was interpreted as phylogenetic relationship by Tullberg (1899), who placed Spalax, Rhizomys, Tachyoryctes, and Myospalax in Spalacidae, an ordering also used by Ognev (1947, 1963a; he did not discuss Tachyoryctes).

Most other classificatory arrangements, implicitly or explicitly, reflect the view that the shared fossorial adaptations indicate evolutionary convergence, not descent from a common ancestor, and have variously combined the genera (also see Topachevskii, 1969, 1976): e.g., Myospalax (Myospalacinae) in Muridae, Spalax and Rhizomys in Spalacidae (Alston, 1876); Rhizomys and Tachyoryctes (Rhizomyinae) and Spalax (Spalacinae) in Spalacidae, Myospalax (Myospalacinae) in Muridae (Thomas, 1896); Rhizomys (Rhizomyinae) and Tachyoryctes (Tachyoryctinae) in Rhizomyidae, Myospalax (Myospalacinae) and Spalax (Spalacinae) in Spalacidae (Miller and Gidley, 1918); Spalax in Spalacidae, Rhizomys and Cannomys in Rhizomyidae, Tachyoryctes (Tachyoryctinae) and Myospalax (Myospalacinae) in Muridae (Ellerman, 1940, 1941); Rhizomys and Cannomys in Rhizomyidae, Myospalacinae and Spalacinae in Spalacidae (G. M. Allen, 1940); Spalax in Spalacidae, Rhizomys and Cannomys in Rhizomyidae, Myospalax as a tribe or subfamily in Cricetidae (Pavlinov et al., 1995a; Simpson, 1945); Myospalacinae and Spalacinae in Cricetidae, Rhizomyinae and Tachyoryctinae in Rhizomyidae (Chaline et al., 1977); Spalacines and Rhizomyines in Spalacidae of Theridomyoidea (Schaub, 1958); Spalacidae and Myospalacidae in Muroidea, Rhizomyidae in Theridomyoidea (Reig, 1980); Spalacinae (Spalax), Rhizomyinae (Cannomys, Rhizomys, Tachyoryctes), and Myospalacinae (Myospalax) in Muridae (Carleton and Musser, 1984; McKenna and Bell, 1997; Musser and Carleton, 1993). These compilations, which recognize three or four separate subfamilies and two or three families, imply independent origin of each, a pattern that echos Ellerman’s (1940:638) pronouncement, "I do not think that there is much doubt that the grouping together of Spalax, Myospalax, Tachyoryctes and Rhizomys is a very unnatural arrangement," and opinions of some fossil rodent specialists: "Spalacidae and Rhizomyidae originate separately from different muroids" (Flynn et al., 1985:608).

Much doubt, nonetheless, is supplied by recent studies in which the cladistic union of Spalax and Rhizomys is supported by the cephalic arterial pattern (Bugge, 1971a, 1985) and by multiple gene-sequence analyses (nuclear LCAT, Michaux and Catzeflis, 2000, and Robinson et al., 1997; LCAT and vWF, Michaux et al., 2001b; nuclear IRBP, DeBry and Sagel, 2001). Further support for uniting Spalax, Myospalax, Rhizomys, and Tachyoryctes as a basal clade relative to other muroids stems from other sequence analyses with wider taxon sampling (Jansa and Weksler, 2004; Norris et al., 2004). These molecular data collectively point to zokors, bamboo rats, blind mole rats, and African mole rats as a monophyletic radiation that is sister-group to all other muroid family-group taxa so far investigated, inferring an early divergence, possibly one of the earliest, from a middle or late Oligocene ancestral stock.

Morphology of fossil molars attributed to spalacines and rhizomyines suggests their common origin. In describing late Miocene African species of Nakalimys and Harasibomys, Mein et al. (2000a:385) allocated them to "Family Rhizomyidae … or Spalacidae," and noted that the early Miocene Spalacine Debruijnia from Turkey (at about 20 million years old, the earliest record among the four subfamiles) "could represent an ancestral form for the African burrowing rodents." Nakalimys was first described from the late Miocene of Kenya and placed in Rhizomyidae (Flynn, 1990; Flynn and Sabatier, 1984), and Pronakalimys from middle Miocene deposits in Kenya is regarded "as the most primitive known African Rhizomyidae" (Tong and Jaeger, 1993:59). The resemblance of Myospalacines to some arvicolines, with their simple occlusal patterns and rootless molars (some extinct species are rooted) impressed Hinton (1926a), but he also noted that the head skeleton and jaw musculature were definitely unlike arvicolines. Lawrence (1991:282) reinforced his observation, remarking that the "character complex myospalacines share with arvicolines is a set of parallel adaptations for propalinal chewing of tough, fibrous plant material," and listing those parallel features as well as synapomorphic traits distinctive to myospalacines and indicative of great phylogenetic distance from arvicolines. Although the molars have different coronal outlines, occlusal patterns in myospalacines are basically similar to those in spalacines and also could have been derived from a form similar to Debruijnia. In turn, the dentition of the Turkish, African, and some Asian Miocene forms recalls occlusal patterns common to Eucricetodon (our observations; Lindsay, 1994, suggested that rhizomyines are closely linked to eucricetodontines), which is known from the late Eocene to early Miocene in Asia, Oligocene in Mediterranean region, and early Oligocene to early Miocene in Europe (Hugueney, 1999; McKenna and Bell, 1997). Hugueney and Mein (1993) wondered if a hypothetical Oligocene ancestor could be unambiguously distinguished from an ancient "cricetid."

A return to Tullberg’s (1899) interpretation of phylogenetic affinities among Myospalax, Spalax, Rhizomys, and Tachyoryctes is the hypothesis best supported by the accumulation of recent morphological and molecular information. This arrangement should be tested by further cladistic analyses of multi-morphological systems and a wider array of genes.