This paper proposes to check previous findings on the correlation of proximate and distal demonstratives in English and Dutch with the accessibility of the intended referent. New data, collected in an experimental setting, lead towards a more finegrained notion of accessibility, that takes temporal factors into account