Horne and Lowe ( 1996) define naming as a fusion of speaker (i.e. tact and
echoic) and listener relations. Within a name relation, all objects that evoke the same speaker-listener behaviour become functionally related to each other, thereby forming a category, whether or not these stimuli bear a physical resemblance to one another. Horne and Lowe's account predicts that the categorisation of physically different stimuli will not occur unless both common speaker and listener relations have been established to sets of such stimuli.
The first study sought to falsify Naming Theory by training common listener
relations, but not common speaker relations, to two potential stimulus classes of
physically dissimilar stimuli, and then testing for categorisation. Of nine participants (age group, 1.5 - 4.5 years), none succeeded in passing the categorisation tests.
However, after subsequently receiving training in speaker relations, and hence
establishing naming, six went on to pass the categorisation tests. Study 2 replicated the methodology of Study 1 with participants from 2.5 - 4.5 years of age. This time, however, rather than establishing only common listener relations between members of each potential stimulus class, the participants also
received concurrent off-task echoic training of the required speaker responses. This investigated whether this training alone, and in the absence of direct training of the common tact responses necessary to complete the full name relation, yield untrained categorisation of physically different objects. Of the 6 participants five categorised the arbitrary stimulus sets into two classes consistent with the experimental names. All five of these participants only categorised when concurrent evidence of the establishment of both listener and tact relations was also shown.
The final study of this thesis trained a common tact response to two potential
stimulus classes. After this training, all three participants (aged between 3.5 and 4.5 years of age) demonstrated successful categorisation and also the corresponding common listener relations.
The findings of all three studies support the view that naming is necessary for
categorisation of physically different stimuli. It is argued that these results cannot be explained by the competing accounts of both Sidman's (1990) Equivalence theory, and Hayes' ( 1986) Relational Frame theory of emergent stimulus classes. It has also been suggested that there exists a relationship between the rapid advances in children's productive repe1toires (the naming explosion) and the onset of exhaustive categorisation (Gopnik & Meltzoff, 1989, 1992). Implications of the current research to this issue is also considered.