Sometimes we argue about cats or about whether there is a largest prime number. Other times we argue about arguments. When we do this, we engage in meta-argument. Most accounts of meta-argument in the literature view it retrospectively: we meta-argue about arguments that have already been made. In so doing, we may find meta-reasons for rejecting an otherwise good argument, among other things. This paper considers meta-argument in the other direction, that is, prospectively. To illustrate this concept, we explore cases where one has meta-reasons for intentionally making bad arguments or where one argues not by offering an argument, but by communicating in a non-argumentative fashion that supports other arguments. We call these later cases para-argument.