In Book Five of the General Theory, Keynes argued, in opposition to his colleague Pigou, that wage and price flexibility, alone, would not suffice to overcome the problem of involuntary unemployment. At this critical point in his overall narrative, Keynes linked together his earlier analysis of liquidity preference and the speculative demand for money, with his discussion of factors influencing both the marginal propensity to consume and the marginal efficiency of capital schedule to suggest that price deflation would have an adverse impact on each of these behavioural relations. In this paper I examine the similarities and differences between Keynesian notions of liquidity preference and recent theories of decision-making under uncertainty aversion, which have allowed researchers to identify the presence of, and distinguish between, both risk-premia and uncertainty-premia in asset markets. Nonetheless I contend that a rigorously Keynesian understanding of liquidity preference can only raise serious doubts about the validity of applying this stochastic control and filtering techniques in a macroeconomic or general equilibrium context. I argue that these doubts have less to do with concerns about agent heterogeneity or coordination failure than with a more fundamental concern about the impossibility of isolating changes in either uncertainty aversion or the magnitude of the relevant stochastic uncertainty constraints as an influence over agent decision-making, from the state variables that appear within the state equation system representing the economic system and its environment.