14 Stanley E. Porter and Matthew Brook O’Donnell subordination, so it is useful to consider whether, rather than needing to conform these lists to traditional criteria, the highly problematic tradi- tional criteria (e.g. is ὅτι a coordinating or subordinating conjunction, as it indicates causality, but uses a verb in the indicative mood?) needs to be adjusted to this research. 5. Conclusion Conjunctions have proved to be a difficult area of study in Greek gram- mar. This is possibly the case because they are both simple and frequent words. It is difficult to read much Greek literature without confronting a number of conjunctions. Most of them are provided with a simple trans- lational equivalent, such as “and”, “but” or the like, and we believe that we have understood their meaning and their use. This clearly is not the case. Like most systems of language, the conjunctions are far more com- plex. We have attempted to differentiate a complex systemic-functional network that describes their complex relations both to discourse and to syntax. We would push for a new system of analysis of the conjunctions, in which specification of a conjunction would involve at least two features: level of conjunction and procedural nature of conjunction. Thus, rather than identifying the conjunction καί and simply translating “and”, or saying “conjunction, meaning and”, we would argue that one should first say that the conjunction is joining units at the, for example, word group level, and indicating continuity between units. One of the results of such analysis would be a clearer understanding of the levels of discourse. An- other would be a move away from seeing translation as the standard by which understanding of a text is determined30. A third would be a clear recognition of the complexity involved in conjoining discourse, and even a step forward in resolving the issue of coordination and subordination. Stanley E. Porter Matthew Brook O’Donnell McMaster Divinity College 1280 Main St. W. Hamilton, ON (CANADA) We believe that translation is a, but certainly not the or the only, means of judging 30 understanding of a text. In fact, often translation hides as much as it reveals. As Gleason so aptly stated, “Translation is a very inadequate means of expressing meanings and must always be used with great caution” (H.A. Gleason, Jr., An Introduction to Descriptive Lin- guistics [rev. ed.; New York 1961] 77).