The target article is a sound, original and well-argued examination of one of the most fundamental topics in the history of anthropological kinship theory. Among its numerous merits, I would single out the much-needed connection the target article establishes between traditional formal and structural analyses, the cognitive anthropology of the 1960s, and more recent cognitive-evolutionary perspectives on the human mind and behavior in general. Many post-Schneiderian kinship studies are characterized by a relativist and social-constructivist perspective. By contrast, the target article clearly posits genealogy as the key constituent of any kinship system – not in terms of a substance (biogenetic connections) but as a structure, a system of relations (see Salazar Reference Salazar, Edwards and Salazar2009). Arguably, though the relationship between the substance of genealogy, human biological reproduction, and its structure is not contingent, the latter can be systematically analyzed independently of the former (cf. Read Reference Read2001). That is the reason kinship terms, as Jones correctly points out, may not have genealogical definitions despite being genealogically structured. I will focus my comment on what I believe is perhaps a rather specific but serious limitation in the way in which Jones' use of Optimality Theory (OT) deals with the Seneca-Iroquois terminologies.
The ranking of constraints, which appears to be one of the main characteristics of this theory, allows for greater flexibility in the identification of the distinctive features of kinship terms than the more conventional componential and reductionist/extensionist approaches. But what I see as problematic is the faithfulness constraint of “distinguish matrikin,” which the author places in the second position in the ranking order. This gives rise to a number of complications; for example, if matrikin need to be distinguished, how can we explain that the Father's Brother's Son, who does not belong to Ego's matrilineage, is classed as a “brother” together with matrilineal relatives such as Mother's Sister's Son? Although Iroquois terminologies have normally been associated with matrilineal societies (despite numerous exceptions including the Yanomamo, Algonquins, Dakota, and Dogon), Lounsbury's (Reference Lounsbury and Hunt1964) reanalysis of Morgan's seminal work (Reference Morgan1871) seems to eliminate the possibility of any significant connection between Iroquois equations and (matrilineal) descent. Stated otherwise, the bifurcate merging characteristic of Iroquois terminologies (merging of Father with Father's Brother and Mother with Mother's Sister) is not connected to the need to single out matrilineally related relatives. Thus I simply fail to see in what way the “distinguish matrikin” constraint makes any improvement on the more traditional cross/parallel distinction. Admittedly, as other authors have pointed out (Allen Reference Allen, Godelier, Trautmann and Tjon Sie Fat1998; Héritier Reference Héritier1981: annex 1), a cross/parallel concept that can be applied equally to relatives of all three central generations – G0 (Ego's generation), G+1 (Ego's parents) and G−1 (Ego's children) – has not yet been formulated. I think this may be the case, as Jones also intimates, because this distinction derives from a more fundamental one. Contrary to Jones's argument, however, this has nothing to do with descent but with what I take to be, following Héritier (Reference Héritier1996, pp. 62–67), the basic binary opposition observable in all kinship terminological systems: that between same-sex and opposite-sex sibling pairs, in virtue of which opposite-sex siblings are never seen as “closer” than same-sex siblings. Héritier contends, correctly in my view, that this fundamental law of kinship originates in the differential value of the sexes as is culturally recognized in all known societies. The four main terminological systems – birfurcate merging (Iroquois and Dravidian, plus Crow-Omaha systems), bifurcate collateral (Latin, Sudanese), lineal (our system) and generational (Hawaiian) – can be all seen as specific developments of the same-sex/opposite-sex sibling relationship. And even the so-called “fifth possibility,” a theoretically possible but sociologically nonexistent terminological system (Hage Reference Hage1997; Héritier Reference Héritier1981, ch. 1), is a direct result of the infringement of that fundamental law of kinship.
It would take me far beyond the scope of the present commentary to elaborate on all the implications of this elementary rule for each terminological system. Here I will focus on the Seneca-Iroquois system, which Jones analyzes as an instance of the workings of OT. In the Iroquois system, children of same-sex siblings (parallel cousins) call each other brother and sister, while children of opposite-sex siblings (cross cousins), are classed simply as cousins. By the same token, children of same-sex (cross) cousins will call each other brother and sister while those of opposite-sex cousins will call each other cousins. In keeping with these simple rules, which clearly derive from the principle according to which same-sex siblings should be seen as closer than opposite-sex siblings, same-sex siblings are “merged” and become the parents of all their children, while opposite-sex siblings are kept separate, becoming “uncles” and “aunts” of each other's children. These rules also account for the fact that the Father's Brother's Son is classed as a “brother” regardless of the matrilineage he happens to belong to, and explain why the Older Mother's Brother's Son's Son is a “cousin” and not a “brother.” Both this individual and Ego are children of opposite-sex cousins: Ego's older mother and her mother's brother's son. The simplicity of this principle contrasts with Jones' argument here, which is unnecessarily convoluted. No further rules, principles or constraints are needed to generate all kin terms of the Iroquois system.
Despite this problem in Jones' argument, his concern with disclosing “fundamental structures of the human mind” through kinship analysis is highly laudable. Further research along these lines will certainly contribute to enhancing anthropology's position among the human sciences by deepening our understanding of the oldest, most universal and most elementary form of human relationship.
You have
Access
The target article is a sound, original and well-argued examination of one of the most fundamental topics in the history of anthropological kinship theory. Among its numerous merits, I would single out the much-needed connection the target article establishes between traditional formal and structural analyses, the cognitive anthropology of the 1960s, and more recent cognitive-evolutionary perspectives on the human mind and behavior in general. Many post-Schneiderian kinship studies are characterized by a relativist and social-constructivist perspective. By contrast, the target article clearly posits genealogy as the key constituent of any kinship system – not in terms of a substance (biogenetic connections) but as a structure, a system of relations (see Salazar Reference Salazar, Edwards and Salazar2009). Arguably, though the relationship between the substance of genealogy, human biological reproduction, and its structure is not contingent, the latter can be systematically analyzed independently of the former (cf. Read Reference Read2001). That is the reason kinship terms, as Jones correctly points out, may not have genealogical definitions despite being genealogically structured. I will focus my comment on what I believe is perhaps a rather specific but serious limitation in the way in which Jones' use of Optimality Theory (OT) deals with the Seneca-Iroquois terminologies.
The ranking of constraints, which appears to be one of the main characteristics of this theory, allows for greater flexibility in the identification of the distinctive features of kinship terms than the more conventional componential and reductionist/extensionist approaches. But what I see as problematic is the faithfulness constraint of “distinguish matrikin,” which the author places in the second position in the ranking order. This gives rise to a number of complications; for example, if matrikin need to be distinguished, how can we explain that the Father's Brother's Son, who does not belong to Ego's matrilineage, is classed as a “brother” together with matrilineal relatives such as Mother's Sister's Son? Although Iroquois terminologies have normally been associated with matrilineal societies (despite numerous exceptions including the Yanomamo, Algonquins, Dakota, and Dogon), Lounsbury's (Reference Lounsbury and Hunt1964) reanalysis of Morgan's seminal work (Reference Morgan1871) seems to eliminate the possibility of any significant connection between Iroquois equations and (matrilineal) descent. Stated otherwise, the bifurcate merging characteristic of Iroquois terminologies (merging of Father with Father's Brother and Mother with Mother's Sister) is not connected to the need to single out matrilineally related relatives. Thus I simply fail to see in what way the “distinguish matrikin” constraint makes any improvement on the more traditional cross/parallel distinction. Admittedly, as other authors have pointed out (Allen Reference Allen, Godelier, Trautmann and Tjon Sie Fat1998; Héritier Reference Héritier1981: annex 1), a cross/parallel concept that can be applied equally to relatives of all three central generations – G0 (Ego's generation), G+1 (Ego's parents) and G−1 (Ego's children) – has not yet been formulated. I think this may be the case, as Jones also intimates, because this distinction derives from a more fundamental one. Contrary to Jones's argument, however, this has nothing to do with descent but with what I take to be, following Héritier (Reference Héritier1996, pp. 62–67), the basic binary opposition observable in all kinship terminological systems: that between same-sex and opposite-sex sibling pairs, in virtue of which opposite-sex siblings are never seen as “closer” than same-sex siblings. Héritier contends, correctly in my view, that this fundamental law of kinship originates in the differential value of the sexes as is culturally recognized in all known societies. The four main terminological systems – birfurcate merging (Iroquois and Dravidian, plus Crow-Omaha systems), bifurcate collateral (Latin, Sudanese), lineal (our system) and generational (Hawaiian) – can be all seen as specific developments of the same-sex/opposite-sex sibling relationship. And even the so-called “fifth possibility,” a theoretically possible but sociologically nonexistent terminological system (Hage Reference Hage1997; Héritier Reference Héritier1981, ch. 1), is a direct result of the infringement of that fundamental law of kinship.
It would take me far beyond the scope of the present commentary to elaborate on all the implications of this elementary rule for each terminological system. Here I will focus on the Seneca-Iroquois system, which Jones analyzes as an instance of the workings of OT. In the Iroquois system, children of same-sex siblings (parallel cousins) call each other brother and sister, while children of opposite-sex siblings (cross cousins), are classed simply as cousins. By the same token, children of same-sex (cross) cousins will call each other brother and sister while those of opposite-sex cousins will call each other cousins. In keeping with these simple rules, which clearly derive from the principle according to which same-sex siblings should be seen as closer than opposite-sex siblings, same-sex siblings are “merged” and become the parents of all their children, while opposite-sex siblings are kept separate, becoming “uncles” and “aunts” of each other's children. These rules also account for the fact that the Father's Brother's Son is classed as a “brother” regardless of the matrilineage he happens to belong to, and explain why the Older Mother's Brother's Son's Son is a “cousin” and not a “brother.” Both this individual and Ego are children of opposite-sex cousins: Ego's older mother and her mother's brother's son. The simplicity of this principle contrasts with Jones' argument here, which is unnecessarily convoluted. No further rules, principles or constraints are needed to generate all kin terms of the Iroquois system.
Despite this problem in Jones' argument, his concern with disclosing “fundamental structures of the human mind” through kinship analysis is highly laudable. Further research along these lines will certainly contribute to enhancing anthropology's position among the human sciences by deepening our understanding of the oldest, most universal and most elementary form of human relationship.