Harm, Affect, and the Moral/Conventional Distinction

Abstract

: The moral/conventional task has been widely used to study the emergence of moral understanding in children and to explore the defi cits in moral understanding in clinical populations. Previous studies have indicated that moral transgressions, particularly those in which a victim is harmed, evoke a signature pattern of responses in the moral/conventional task: they are judged to be serious, generalizable and not authority dependent. Moreover, this signature pattern is held to be pan-cultural and to emerge early in development. However, almost all the evidence for these claims comes from studies using harmful transgressions of the sort that primary school children might commit in the schoolyard. In a study conducted on the Internet, we used a much wider range of harm transgressions, and found that they do not evoke the signature pattern of responses found in studies using only schoolyard transgressions. Paralleling other recent work, our study provides preliminary grounds for skepticism regarding many conclusions drawn from earlier research using the moral/conventional task.

Varying versions of moral relativism: the philosophy and psychology of normative relativism

“…Given the previously stated critique (Section 4.1.1) that it should be up to the participant to explicitly classify rules as 'moral' or 'conventional,' there are not sufficient grounds to conclude that only conventional rules can become relative, while moral rules cannot. Kelly et al (2007) provide findings consonant with the suggestion that moral rules can be thought of as relative. They find that participants are indeed more likely to say that more historically and locally variable moral rules against slavery or cannibalism are ok or not depending on time and place.…”

“…Studies that included a wider range of scenarios and did not have inclusion or characterization criteria based on Turiel's (1983) classification did not find this clear-cut conceptual distinction (e.g. Huebner et al 2010;Nichols 2004;Kelly et al 2007). Finally, there are cultural differences in how people classify transgressions.…”

The case of the drunken sailor: On the generalisable wrongness of harmful transgressions

“…We adapt Kelly et al's (2007) ''whipping generalisability'' scenario by explicitly varying one versus two aspects of convention, while avoiding the introduction of HJR confounds. We also add a justification question to check for confounds.…”

“…We adapted the ''whipping generalisability'' scenario (see Kelly et al, 2007) such that, in the ''high'' condition, two aspects of convention were explicit, while in the ''low'' condition only one aspect was explicit; to avoid HJR confounds all other features remained the same across conditions. Each participant read two scenarios.…”

“…Critics have questioned the existence, exact interpretation (see Discussion section), and appropriate experimental tests of the M/C distinction (Kelly, Stich, Haley, Eng, & Fessler, 2007;Sousa, 2009;Sousa, Holbrook, & Piazza, 2009;Stich, Fessler, & Kelly, 2009). Here we raise methodological concerns regarding the measurement of generalisability.…”