Here we have a 3d scatterplot representing the spread of a set of 400 abstract and concrete English nouns within a multi-dimensional space bounded by three broad cognitive domains (magnitude, sensation, emotion). Abstract words are reflected by the black bubbles, and concrete words are the red bubbles. Their positions in this x,y,z plane reflect ratings from almost 400 adults from Amazon's Mechanical Turk. It's interesting -- Abstract and concrete words cluster in unique ways in this space. We're arguing that a multi-dimensional 'topography' approach like this obviates the need for multiple semantic systems (e.g., verbal for abstract words, visual for concrete words). The other neat thing about this approach is that deficits in concrete or abstract word representation can be modeled in terms of dimensionality reduction (e.g., a sensory impairment might reduce the dimensions to only magnitude and emotion). We just submitted this article for review. Its recent ancestor was just also published in Frontiers in Human Neuroscience. Here's the R code for generating this 3d scatterplot using the rgl package: