Fascinating book, highly recommended by a Goodreads friend. It's amazing to see how far the science of meaning has progressed since I was last studying it some 35 years ago. The book is mostly about "embodied simulation"--the process by which we "make meaning" (more on that in a second) by imagining ourselves doing actions which words describe. I'm astonished at the ingenuity with which experimenters have been able to design studies to tease out various nuances in this view. But I'm an old guy, and I can't rid myself of the old-fashioned view that a speaker has an idea, he chooses the words which convey it, the hearer "understands" the words and receives the meaning. The hearer doesn't "make" the meaning, he "understands" it. (Bergen uses these words more-or-less interchangeably.) Luckily for me, in a closing chapter Bergen finally admits that my view can't be wholly wrong; but it certainly has nothing to do with the main argument of his book. I also have a little trouble with his apparent assumption that computers might be designed to do the same thing we do when we understand each other. I'm very dubious about that, if for no other reason than John Searle's thought experiment involving bits of paper with Chinese characters on them passed through a slot in a wall to a person who understands no Chinese, but who has a book of instructions on how to respond to any combination of characters. The person follows instructions, writes what the book tells him to, and when his written response is received, communication appears to have taken place. I would say that the conveyance of meaning has been simulated. That, for me, is computer communication in a nutshell. Bergen mentions this thought experiment in a different context, but he doesn't draw from it what I do. Still, a fascinating book, clearly written, with excellent examples, and close reasoning. I'm gratified that so many people this smart are working on these questions.