Synchronization and a binding dimension handle segmentation.

• Segmentation can involve dividing a scene into separate objects, largely on the basis of contiguity.

  In a visual context:

  

  In a linguistic context:

  

• Segmentation can involve grouping sets of objects, largely on the basis of similarity.

  In a visual context:

  

  In a linguistic context:

  

• In the patterns in the Marcus et al. experiment, the babies are apparently grouping the words by similarity.

• Handling segmentation (of either type) requires a means of marking different elements as belonging temporarily together, as being "the same."

• One popular way of doing this is with a separate binding dimension in addition to activation along which units vary. Units which coincide along this dimension represent the "same thing".

  In visual contexts:

  

  ---

  

  In linguistic contexts:

  

  ---