The notion of acceptability has played a crucial role in linguistics. Formal sentence acceptability experiments are relatively recent, but standardly make use of a factorial design, multiple lexicalizations of the stimuli, full counterbalancing of the stimuli, well-designed filler items, and an appropriate response method. Such experiments are sensitive to grammaticality, of course, but also to the presence of a dependency, the length of the dependency, and the frequency of the lexical items and structures. These experiments are useful for testing claims of (un)acceptability, but also for making cross-linguistic comparisons and comparing populations of speakers. Formal acceptability experiments are similar to traditional methods of collecting acceptability judgments, but each can do things that the other cannot and both have important roles to play in syntactic research.

The starting point of my book is straightforward: military coups have been the main causes for autocrats’ downfall globally throughout the twentieth century. Consequently, the survival of nondemocratic leaders in coup-prone countries is above anything else a function of successful coup-proofing. I show how frequent coups have been in developing countries following decolonization, but also, how harsh the fate of fallen leaders have typically been following military takeovers – particularly in the Arab world. To coup-proof, autocrats used ideational factors (i.e., fostering shared aversions); or material factors (i.e., counterbalancing; promoting the material interests of senior officers; divide-and-rule tactics); or combinations of both. I show how coup-proofing tactics structure civil–military relations differently, but also, amplify or reduce vertical and horizontal rifts in officer corps. I then link these variations to military behavior when popular uprisings challenge the authoritarian status quo.

This chapter introduces the basic principles of experimental design, covering fundamental control groups,such as approaches of control for subject loss or for order of treatments.