1. 5
    1. 4

      I think they are approaching the problem at too abstract a level. And, no, I’m not being facetious.

    2. 2

      I have investigated this sentence: “The first author’s analysis of political metaphors in the Java libraries (Blackwell 2006) has already drawn attention to the fact that social and political concerns may be central to future research in the psychology of programming.”

      There is a confusion in reference, but (2006a) “Metaphors we program by: Space, action and society in Java.“ was a very interesting read. It did not conclude anything on politics, but did a literal corpus analysis of Javadoc texts. Many surprising concepts were revealed, but suggestions or convincing explanation lacking. Examples:

      • Economy: “deliver”, (“return”, “import” were disqualified due to keywords of Java)
      • Law: “contract”, “forbidden”, “obliged”, “constraint”
      • Biological: “head”, “tail”
      • et cetera
    3. 2

      Mistaking the map for the territory can be an issue, but that does not mean, as this paper seems to imply, that the abstraction of the map is an issue in itself.

      We hope to explore an alternative style, in which professional “good practices” are experimentally set aside, allowing end-user programmers to construct programs that are not reliant on abstraction or other central features of computational thinking. Rather than changing the user, assisting him or her to think more computationally, we ask whether computers can be made more accessible to those who, for whatever reason, prefer not to do so.

      By necessity, we all use abstraction to deal with a complex world. I really can’t see how pushing the formulation of that abstraction onto end users will help anything. Or how, practically, you’d go about implementing a a system that’s not reliant on abstraction yet is still useful.