This study is devoted to some synchronic aspects of the sDm.t=f verb form, primarily its meaning and uses in Classical Egyptian. In the introduction some attention is paid to the history of the studies of the form and its origin, an aspect which will receive no further consideration. In accordance with present common opinion the sDm.t=f is here considered to belong to the suffix conjugation. Ch. I is primarily concerned with the active Dr sDm.t=f construction, but begins with a general introduction to the form. The author proceeds from the description in Gardiner's standard grammar. Three uses are distinguished there: (1) the negative construction n sDm.t=f "before he has/had heard", "while he has/had not (yet) heard", "he has not (yet) heard" (a remark that the temporal notion "not yet", "before" is only very rarely absent is added); (2) the use after the prepositions/conjunctions m "when", xft "when", m-xt "after", mi "like", r "until", and Dr "since", "from the moment that", but also "before", "until"; (3) the use as a narrative tense. Gardiner notes that the use after the first four prepositions/conjunctions is not certain. It might concern the infinitive with -t-ending, since only examples of verbs with a feminine infinitive are known. Next, the more recent opinions on the uses and meaning of the sDm.t=f are described. In accordance with common opinion, the sDm.t=f in narrative use is considered non-existent, because all forms used in evidence may also be regarded as infinitives. While there is unanimity of opinion on the existence of n sDm.t=f, this is not so for the use after prepositions. Some scholars recognize the use of the form after all six prepositions/conjunctions mentioned by Gardiner, but others only the two certain uses after r and Dr. Two problems regarding r and Dr confront us here. The first is that some scholars also claim the meaning "so that he hears/heard" (has/had heard)" for r sDm.t=f, in addition to the generally accepted "until he hears/heard (has/had heard)"; this is only a minor problem, however, since both translations have the same time reference. A weightier problem for the author is that for the indubitable construction Dr sDm.t=f some scholars accept both a relative past translation with "since/from the moment that he has/had heard", and a relative future one with "before he has/had heard", whereas others accept only one or other of the two. The author does not believe it to be coincidence that only the indubitable constructions n sDm.t=f and r sDm.t=f can be translated with the conjunctions "before" and "until" expressing the relative future, respectively. Relating this relative future time reference to the uncertain use after the prepositions/conjunctions m, m-xt, mi, and xft, which do not have this time reference, he feels justified in leaving them out of consideration. Westendorf's hypothesis that the sDm.t=f verb form denotes the relative future is taken up again.

On this basis the provisional hypothesis that the three indubitable sDm.t=f constructions (and not the form) share a relative future time reference is formulated. Then, the only problematical construction in this respect, Dr sDm.t=f, requires closer observation. Before proceeding to this construction, the basics concerning the concepts "tense" and "time reference", and concerning the distinction between relative and absolute tenses or time reference as used in this study are described. It is pointed out that the absolute time reference of present, past and future is related to the moment of speaking, whereas this is different with relative time reference. In this case there is a Reference point, that is, some point in time given by the context. This is not necessarily the absolute present, which is, however, always available as the Reference point. Temporal clauses expressing the notions "while/when (simultaneously)", or "after/since"/ "while/when already", or "before/until"/ "while yet to" have relative present, relative past and relative future time references, respectively. A finite relative tense grammaticizes as relative present tense simultaneity ("I hear/am hearing") relative to a Reference point in the context; as relative past tense anteriority ("I have already heard") relative to it; and as relative future posteriority ("I am yet to hear") relative to it. The relative tense may be related to the type of Reference point which itself is an 'empty' anchor point, thus not having absolute time reference of its own (in contrast to, for example, absolute time adverbials like "tomorrow", "yesterday", "today"). Therefore, a distinction is now made between the Reference situation and the Reference point as a word (or expression) to which the relative tense is syntactically related and which relates itself by its relative temporal meaning to the Reference situation providing the absolute time. Of course, the Reference point and the Reference situation may well coincide. For the sake of clarity it should be pointed out that our 'modern languages' do not possess finite relative tenses, but they do have verb forms that combine absolute and relative time reference, such as the pluperfect and the future perfect, both of which express anteriority relative to a Reference point, which in the case of the pluperfect is situated before the absolute present, but in the absolute future when it concerns the future perfect. Of course, the reference time and the time of the wider context may differ, for example, "yesterday he said: 'I shall wash myself before I enter the temple hall'". Entering the temple is located in the future relative to the absolute future Reference situation of "going to wash myself", while the time reference of the context is absolute past: "yesterday he said". This exposition on time reference is the more necessary since Gardiner, and others after him, have stated that the time of the sDm.t=f in all its certain uses is the relative past. This bluntly contradicts the translation of precisely those certain constructions r sDm.t=f with "until" and n sDm.t=f/Dr sDm.t=f with "before". Among the certain uses the statement is only in agreement with the translation of Dr sDm.t=f with "since/from the moment that". This construction is now concentrated upon. The discussion of the evidence shows that "before" is an obligatory, or at least a suitable, translation of Dr sDm.t=f, not only in the few indubitable examples--that is to say, where the form cannot also be the infinitive--but also where we are dealing with morphologically ambiguous forms which can, nevertheless, be identified with certainty as the sDm.t=f on other grounds. It is postulated that all examples of Dr + alleged sDm.t=f with the meaning "since/from the moment that" contain the feminine infinitive (see further Appendix .

Next, the equivalent of the core meaning of the substantive Dr(w) "limit" with respect to the derived Dr as temporal preposition/conjunction is defined as "(at) an ultimate time point". The translation with "before" of Dr sDm.t=f can be brought into agreement with the common role of 'starting point' ("since from the moment that/of") of Dr as a preposition/conjunction in Dr + Prospective sDm=f/ + infinitive/ + substantive if it is assumed that in these cases "since/from the moment that/of" is only an interpretation, as in English "(since/from) the moment that he was born/of his birth he has appeared as king". Also, in Dr sDm.t=f the time point represented by Dr can be regarded as located at the beginning of the part of the time line occupied by the situations of the main and the subordinate clauses: ia=i wi Dr aq.t=i r iwnyt actually means "I shall wash myself at a time that I am to enter the temple hall yet". The hypothesis is now narrowed down to the formulation that the sDm.t=f verb form is a relative future tense meaning "he is yet to hear relative to a Reference point/situation in the context". This would explain the location of the Event in the subordinate clause in a future relative to the 'empty' Reference point of Dr, which coincides with the Reference situation in the main clause and forms the transfer point for the reference time. In ch. II, on the active r sDm.t=f construction, the author proceeds from the results obtained in ch. I for the sister construction Dr sDm.t=f. It is postulated that r sDm.t=f has the exclusive temporal meaning "until he hears/heard (has/had heard)", in contrast to r + Prospective sDm=f with consecutive (/final) "so that he hears/heard". Attention is drawn to r + infinitive in the Pseudoverbal Sentence iw=f r sDm, which always has the temporal significance of "underway in time to" ® "shall, will". Generally speaking, however, the construction as adverbial adjunct is thought to have the final meaning "to the purpose of", but the author assumes that in the latter use too the meaning can be temporal "until". Furthermore, it is emphasized that identification of a verb form + expressed agent as a suffix conjugation form and not as the infinitive only rests on an assumption. In this view r + infinitive as adverbial adjunct has a semantically intermediate position which allows the translation with temporal "until" as well as with modal "to the purpose of".

After establishing that the r sDm.t=f construction occurs while related to the three possible absolute reference times, the author discusses the constructions r snb.t=f "until he is better" versus r snb=f (Hr-awy), which frequently occur in the medical texts. The latter construction must be translated either with "so that he will be better (very soon)" when a Prospective sDm=f is identified, or with "until his being better" (r snb=f) / "to the purpose of his being better (very soon)" (r snb=f (Hr-awy)) in case of identification as the infinitive. Although some scholars are of the opinion that the construction r Hst=f, known in a number of variants, contains the sDm.t=f and must be translated with "so that he praises/praised", the author, following Gunn, identifies a relative form: r Hst=f "with a view at what he praises/praised". To corroborate the postulate that r + Prospective sDm=f has only modal and not temporal meaning, some examples are presented of r + forms without -t-affix as adverbial adjunct which favour an identification as masculine infinitive, with the possible translation "until". Finally, in this chapter the roles of the preposition/conjunction r and the sDm.t=f are subjected to closer observation. It is demonstrated that the preposition/conjunction r lengthens the main clause situation into a process which continues "all the time before" a final point expressed by the sDm.t=f and the beginning of which serves as reference time. The hypothesis of the sDm.t=f as being a relative future tense in this construction, too, can be maintained without problem. In ch. III, on the active n sDm.t=f construction, the point of departure is that the form is a unity, and therefore an attempt must be made to bring the three uses n sDm.t=f and Dr/r sDm.t=f under one common semantic denominator. In the section establishing that, just like the r sDm.t=f construction, n sDm.t=f also occurs as related to the three possible absolute reference times, the common translation "before he has/had heard" is provisionally employed, in order to underscore that in this translation the construction can be interpreted in terms of relative future time reference and that the rather common acceptance of a relative past time reference is incorrect. Next, a linguistic analysis by Comrie of the 'still', 'no longer' and 'not yet' tenses in a Bantu language is quoted in some detail. Comrie points out there that a presupposition (i.e. something underlying the utterance and taken for granted by the speaker, and assumedly by the adressee as well) may underlie the assertion of a verb form, and is then grammaticized in its meaning. Furthermore, he explains that, although it seems that in an utterance like "he has not yet heard" it is asserted that "hearing" will occur in the future, this time reference is only an implicature, i.e. something that can reasonably be inferred, though it is not actually said. According to him, the utterance asserts no more than "he has not heard in the past up to the present". Furthermore, he argues that the distinction between a presupposition and an implicature is that a presupposition cannot be cancelled when disclosed under negation, which can thus contradict only part of the meaning; in contrast, it is possible to cancel an implicature with arguments in the context. Implicature is not part of the meaning of a verb form, in contrast to presupposition as underlying the assertion. Since in Dr sDm.t=f specifically the original meaning of Dr as being neutral to absolute time reference proves that the verb form has relative future meaning, this time reference must be part of the meaning of the verb form. On the one hand, relative future time reference as implicature in Egyptian n sDm.t=f must therefore be excluded, while on the other Comrie has stated that an implicature is present in its most suitable translation "(while) he has not yet heard". Because this translation cannot be the literal one, we must be dealing with a 'translation trap'. When the n sDm.t=f construction is used in a situation which is located in the absolute past, it becomes evident that it is not being asserted that the Event has not taken place, but only that this Event has not occurred at the time of the Reference situation. The occurrence of the Event in a relative future is not affected by the negation. This seems to me to reveal a presupposition of an 'Event in the relative future'.

Since the relative future time reference is part of the meaning of the sDm.t=f, adaptation of the usual translation "while he has/had not yet heard" is required. In a literal translation the sentence sr=sn Da n iy.t=f with absolute past time reference must be rendered "they foretold a storm, when 'it was yet to come' was not holding", and with absolute present time reference "they foretell a storm, while 'it is yet to come' is not holding". The view that the sDm.t=f is a relative tense offers a solution to the question of how it is possible that the negation in n sDm.t=f does not deny the occurrence of the Event in a relative future, but only its occurrence at the time of the Reference situation. This is because such a tense grammaticizes two time points in its meaning, one of the Event, the other of the Reference point/situation. When thus, by the very meaning of a relative tense, the Event is related in time to a Reference point/situation which is grammaticized in that meaning, that Event is presupposed, and as such a presupposition. It is pointless to relate a nonexistent or potentially nonexistent Event to a Reference point/situation. In n sDm.t=f this presupposition of a 'relative future Event' is revealed: being part of the meaning of the verb form, the presupposition cannot be denied by the negation, whence the negation can only turn to the non-presuppositional part of the meaning and assert that the occurrence of the Event does not take place at the time of the Reference point/situation. This all results in an obligatory translation with "not yet", but for all practical purposes the usual and suitable translation with a past tense should be maintained. Two cases of n sDm.t=f constructions--whether correctly identified as such or not--where the 'not yet' notion seems to be absent are briefly discussed. 
The rare indubitable examples of n sDm.t=f in the main clause are presented, including--in anticipation--some passive sDm.t=f forms which form important evidence for this use. Finally, the omission of the suffix pronoun in certain n sDm.t=f constructions is studied, which phenomenon primarily occurs when it concerns n sDm.t=f in the 'true' and the 'virtual' relative clause. This omission can be explained from the specialization of the meaning of n sDm.t=f to indicate something like "before the Event", and from the situation that the negation and the verb form are bound together. Ch. IV deals with the passive form sDm.t=f/ms.(y)t=f (where the verb msi represents the ult.inf. verbs, in which an optionally defectively written -y-infix occurs in front of the t). First, it is emphasized that this form is simply the finite passive form of the active sDm.t=f, and not, in a more detached formulation, its 'passive counterpart'. Attention is drawn to the formal resemblance between the passive sDm.t=f/ms.(y)t=f and the defectively written passivized sDm.t(w)=f. Since this has an effect on the translation, a case is extensively discussed in this light. Then, the author presents a series of passages from the religious literature which contain passive n sDm.t=f/ ms.yt/ms.(y)t=f constructions. They deal with the conception, the birth and the existence of the deceased speaker in primordial times, before the gods etc. had come into existence. After the discussion of two certain examples of n fx.t=f "when he had not yet been circumcised" and of the rare other indubitable n sDm.t=f constructions with passive meaning, instances of the passive r sDm.t=f/r ms.(y)t=f are studied, as well as one with passive Dr ms.(y)t=f. Now that all obstacles to the definition of the sDm.t=f in its three uses as a relative future tense have been removed, Ch. V turns to other aspects of the verb form proper and the relationship with other relative tenses in Classical Egyptian. The chapter begins with a survey of the forms of the stem in the mutable verbs. Next, it is pointed out that the relative future tense sDm.t=f can best be studied from the viewpoint of the existence of other relative tenses in Classical Egyptian, the Circumstantial sDm=f and sDm.n=f, which are the relative present and the relative past tense, respectively. Their characters as relative tenses are discussed, as well as the problem of their adverbial nature in both the circumstantial clause and in the main clause constructions iw sDm=f/sDm.n=f and aHa.n sDm.n=f. Whereas the Circumstantial sDm=f/sDm.n=f in the subordinate clause are usually related to a reference situation in the main clause, the main clause constructions iw sDm=f/sDm.n=f and aHa.n sDm.n=f have their particles as absolute present Reference point and, taken in their entirety as compound verb forms, they have absolute time reference themselves.

The semantic function of the Circumstantial sDm=f/sDm.n=f as relative tenses is identical in the main and the subordinate clauses. For this reason the author takes a stand against Collier's recent analysis of the Circumstantial sDm=f/sDm.n=f after particles as verbal verb forms in the sense of Polotsky's Standard Theory. Since the verb form and the particle are bound, semantically as well as syntactically, the compound verb forms are, in the view of the author and others, best viewed as verbal verb forms in their own right. However, Collier's view that the morphologically adverbial forms in iw sDm=f/sDm.n=f/substantive and aHa.n sDm.n=f/substantive cannot have an adverbial function is accepted. In a surface analysis the constructions display an order of Predicate - Subject. Given the absence here of the well-known adverbial order of Subject - Predicate, they do not fit into the transpositional analysis of Polotsky. 
The phenomenon can be explained from the regular order of Predicate - (substantive as) Subject with the suffix conjugation forms and from the origin of iw sDm=f/sDm.n=f and aHa.n sDm.n=f in an Adverbial Model, which is still active in Middle Egyptian in iw=f/substantive + Circumstantial sDm=f. Next, the presupposition type of 'relative time location of an Event' which is at the basis of the relative tense sDm.t=f is connected with the sister forms Circumstantial sDm=f/sDm.n=f. This may explain why the negative counterparts of these forms, for the purpose of negation of the Event in the main clause with iw/aHa.n as well as in the subordinate clause, are n + Indicative sDm=f/sDm.n=f (or Substantival sDm.n=f), i.e. verb forms which as absolute tenses are not subjected to this presupposition. Arguments are brought forward concerning why the sDm.t=f form viewed as a unity cannot be an adverbial verb form. If so, the sDm.t=f should be able to function independently in the (or, as) temporal clause, just as the Circumstantial sDm=f/sDm.n=f do. The construction with r and Dr proves that this is not so. Also, *iw sDm.t=f in the main clause is impossible.

To determine the verb form category, the author refers to an earlier article in which he attempted to demonstrate that not only the construction n sDm=f (is), but also n sDm.n=f (is) contains an 'Indicative', verbal verb form, this in contrast to the view of Polotsky and his followers who identify the Substantival sDm.n=f in n sDm.n=f (is). Polotsky's view concerning a syntactic connection between n sDm.n=f and n sDm.t=f is now applied to promote the suggestion that n sDm.t=f also contains a verbal verb form. Assuming that the sDm.t=f covers only one form, a verbal verb form in all three uses does not per se contradict the fact that after the prepositions/conjunctions r and Dr the function is 'but' substantival--instead of the form as substantival--, in view of the fact that such a multifunctionality is also known from, for example, the Stative 1st person singular, the Passive sDm=f and the Indicative sDm=f (with respect to the first two forms, of course, only in so far as they are viewed as a unity). In other words, when a verb form exercises one of the functions of the parts of speech other than the verbal, this does not automatically entail transposition into an adverbial or substantival verb form. If the sDm.t=f form is thus indeed a genuinely verbal verb form, then n sDm.t=f must on principle be a main clause construction, like n sDm=f/sDm.n=f, though in practice it almost always functions circumstantially. The few instances of n sDm.t=f in the main clause have already been assembled in ch. III. The practically always circumstantial use of n sDm.t=f and its '(while) not yet' notion invite a short confrontation with the main clause constructions nn sDm=f and iw=f r sDm. The negative construction nn sDm=f differs from n sDm.t=f in that it asserts that an absolute future Event will not take place. The absolute future iw=f r sDm is the main clause counterpart of n sDm.t=f. Generally speaking, the two constructions have a complementary distribution. On principle there is no syntactic obstacle at all to iw=f r sDm exercising a circumstantial function, just as iw=f Hr/m sDm can. This use of iw=f r sDm as restricted to the main clause may explain why in later Middle Egyptian and Late Egyptian iw has become a bound constituent only in this Pseudoverbal construction. Negative utterances usually take up a matter mentioned before in the context, and the combination in n sDm.t=f of relative tense and negation predisposes the construction for use as a negative clause subordinate to a main clause; the exceptional use of n sDm.t=f in the main clause can be explained by the fact that this need not be so per se, thus, for example, in dialogues or correspondence. After it has been pointed out that iw sDm=f/sDm.n=f and aHa.n sDm.n=f are, to some extent, syntactically comparable with n sDm.t=f, namely with respect to the dependence of their verb forms on the Reference points iw/aHa.n and n, respectively, the question arises of why, if the sDm.t=f is a verbal verb form, it cannot operate independently as a relative tense and, instead, n sDm.t=f occurs. After all, the main clause is typically the environment where verbal verb forms occur. 
On account of the presence of a Reference point in iw sDm=f/sDm.n=f and aHa.n sDm.n=f in the form of the particle, and in n sDm.t=f in main clause use in the form of the negation, it is concluded that in the Classical Egyptian verbal system relative tenses can only function in the main clause if the Reference point is also represented there. The negation is not a Reference point with absolute time reference, but rather functions as syntactic transfer point establishing the time coincidence with the Reference situation. The next question posed is whether the sDm.t=f grammaticizes the perfective aspect, according to a frequent characterization of the verb form. The t-morpheme in the form is judged by the author to mark the future, just as in other future verb forms with -ti/y-affix (see below). The fact that the sDm.t=f locates an Event in a future relative to a Reference point permits of perfective meaning as deriving from this to be spoken of, but does not ascertain that the form has a perfective aspect. So far, the presupposition of a 'relative future Event' has only played a part when grammaticized in the relative future tense sDm.t=f, but now it is connected with the speaker and his motivation for using the sDm.t=f in an utterance. When the sDm.t=f is related to a reference situation in the absolute past, the very fact that the occurrence of the Event has usually meanwhile been realized confirms the truthfulness of the speaker's presupposition.

But when the Reference situation is not located in the absolute past, objective factual confirmation that the Event has been realized is impossible. Nevertheless, the speaker now also proceeds in his utterance from the realisation of the Event in the absolute future, although he may not even be able to exercise control over this through his own actions. Some instances of n sDm.t=f where the Event has been 'aborted' by the action in the Reference situation--among others, the famous passage on the coregency in the Instruction of Amenemhat I-- seem to contradict the presupposition of a 'relative future Event'. Why would a speaker for reference to an irreal Event--of which at best he may find that it should have occurred--use a construction which grammaticizes, as a presupposition, the expectation that this Event would occur in a future located relative to a reference time? On the one hand, the use of the construction under this condition may be explained by the fact that in other situations involving the irrealis of the past, too, Classical Egyptian uses a verb form, the sDm.n=f, which normally expresses a fact that has taken place; also, the negative construction n sDm=f may be employed to express the irrealis of the past. But on the other hand it may be that in n sDm.t=f an Event is simply located in a future relative to a Reference time, without being concerned about the reality of the occurrence of the Event. This chapter concludes with some remarks on the history of the sDm.t=f and on the other verb forms in which a morpheme written t or ti/y marks the future. Appendix A deals, among other things, with the nomen actionis ending -t of verbs with a masculine infinitive. In Appendix C it is argued that the 'not yet' tense in a Bantu language, which has been analysed by Comrie as only involving the implicature of a future Event, must be analysed in the same way as n sDm.t=f, i.e. in terms of a presupposition. Also, the English expression "not yet" as a unity can be analysed as containing this presupposition. The last chapter, ch. VI, is devoted to the ti-morpheme which when written as t occurs in the sDm.t=f and the Subjunctive Prospective sDm=f forms iw.t=f/in.t=f, and as infix written as t(i/y) in the future participle sDm.ty.fy. Presumably the writing with i/y indicates a vowel. Gunn's assumption that the basis of the sDm.ty.fy is an active Future Participle sDm.ti which is invariable is followed. The latter point would explain the unique use of the suffix pronouns to indicate the number and gender of the antecedent. After presenting the evidence for this form with both masculine and feminine antecedents, Gunn's derivation of this Future Participle from a feminine participial form is rejected as too improbable, particularly now that the existence of a ti-morpheme marking the future has been demonstrated. The Future Participle sDm.ti and the sDm.ty.fy display a number of common features: both are active, facultatively write a -w-infix in the case of ult.inf. verbs, and have the same stem (except with the anomalous verb wnn). The author claims that these forms are the participial counterparts of the suffix conjugation form sDm.t=f, a claim supported by the agreement in the stems of the forms. It may be that this ti-morpheme is a relative of the infix -ta- in the Akkadian Perfect iptaras that is particularly used when indication of an Event as posterior to another one is necessary. The Prospective Participle that is almost exclusively used in the passive voice and has the endings -y and -ti in the masculine and feminine, respectively, is regarded as unconnected with the forms having the future morpheme ti, but related to the Prospective sDm=f, which in Middle Egyptian may show a future morpheme y. Here, too, there is a correspondence between the stems of the (passive) Prospective Participle and the Prospective sDm=f. The circumstance that the writing of the ending -ti in the feminine (passive) Prospective Participle has been able to maintain itself, may well have been positively influenced by the existence of the future morpheme ti. Evidently, these two future/prospective participles have a complementary active/passive distribution. A bibliography and an index of text citations are added.