Requirements elicitation research is reviewed using a framework categorising the relative ‘knowness’ of requirements specification and Common Ground discourse theory. The main contribution of this survey is to review requirements elicitation from the perspective of this framework and propose a road map of research to tackle outstanding elicitation problems involving tacit knowledge. Elicitation techniques (interviews, scenarios, prototypes, etc.) are investigated, followed by representations, models and support tools. The survey results suggest that elicitation techniques appear to be relatively mature, although new areas of creative requirements are emerging. Representations and models are also well established although there is potential for more sophisticated modelling of domain knowledge. While model-checking tools continue to become more elaborate, more growth is apparent in NL tools such as text mining and IR which help to categorize and disambiguate requirements. Social collaboration support is a relatively new area that facilitates categorisation, prioritisation and matching collections of requirements for product line versions. A road map for future requirements elicitation research is proposed investigating the prospects for techniques, models and tools in green-field domains where few solutions exist, contrasted with brown-field domains where collections of requirements and products already exist. The paper concludes with remarks on the possibility of elicitation tackling the most difficult question of ‘unknown unknown’ requirements.