Hermeneutical Factors in Determining The Beginning of the Seventy Weeks (Daniel 9:25)

by Vern Sheridan Poythress

Westminister Theological Seminary

[Published in the Trinity Journal N.S. (1985) 131-149. pdf. Used with permission.]

What is the date for the beginning point (terminus a quo) for the period of “seventy weeks” prophesied in Dan 9:24–27? There has been much debate about this. But sufficient attention has not yet been paid to the role of hermeneutical differences in reaching a conclusion. One crucial factor is what we mean by “grammatical-historical interpretation,” when applied to Dan 9:24–27.

1. Alternative Dates

First of all, let us look at the major alternatives. According to Dan 9:25, the “seventy weeks” begins with a “decree to restore and rebuild Jerusalem” (Dan 9:25, NIV). What is this “decree”? Most conservative interpreters have identified the decree either as the decree of Cyrus king of Persia in about 538 B.C. (Ezra 1:2–4),¹ or the decree of Artaxerxes concerning Nehemiah’s rebuilding (Neh 2:8–9), in 445 or 444 B.C.² Other dates have sometimes been proposed (e.g., 457 B.C.),³ but I wish at this point to con-

TrinJ 6:2 (Fall 1985) p. 132

centrate on the main options.

The date of 444 B.C. is particularly attractive because it is almost exactly 69 x 7 = 483 years before the death of Christ. Now Dan 9:26, according to many, predicts the death of Christ after 69 weeks. Reckoning backwards to the beginning of the weeks, with each week seven years long, leads to a date in Nehemiah’s time. However, such a procedure appeals to facts not known when Daniel was originally written. Are we justified in doing this? Is this a part of grammatical-historical interpretation? Grammatical-historical exegesis, in the narrowest sense, interprets each passage against the background of its original historical context, taking into account what information is known and not known at the time. Hence we must consider how the principles of grammatical-historical interpretation impinge on the interpretation of this passage.

2. Hermeneutical alternatives

There are at least three alternative hermeneutical approaches. First, we might argue that we should confine ourselves to grammatical-historical interpretation in the narrowest possible sense. We should determine only what the passage expressed in its original context. We should rigorously confine ourselves to what people could be expected to know in Daniel’s lifetime. All appeals to dates long after Daniel’s time are illegitimate. Even if Daniel by a special work of the Spirit was given more information about the future, only what is expressed in the text, that is, what is potentially available to the ordinary godly reader, is part of the record. Let us call this approach “narrow” interpretation.

Second, we might argue that Daniel 9 should simply be interpreted in the light of the entire canon of the Bible, or perhaps even in the light of whatever extra-biblical information that we can obtain about fulfillment. No particular interest need attach to the limitations of the original circumstances of writing. Let us call this approach “canonical” interpretation.

Third, we might argue that biblical interpretation, properly understood, includes more than the “narrow” approach. It should reckon with the original situation and the limitations of knowledge at the time; but especially in the case of prophecy, it should also take into account what we know of later fulfillment.

Each of these approaches offers attraction and possible liabilities. I prefer the second, “middle-of-road” approach. It maintains the importance of the original situation on the one hand, and on the other it allows for the importance of progressive revelation and later additions to the canon.

Suppose, then, that we choose the middle way. Do we just mix together, willy-nilly, the “narrow” form of grammatical-historical exegesis with the “canonical” approach? This could mean that we simply appeal to whichever approach justifies our favorite views. We need more rigor than this. One way of introducing rigor would be to say that both approaches (narrow and canonical) must be applied to every text. In order to have historical control, we ask that the canonical approach build on, not undermine, the results of the “narrow” approach.

Hence, I believe that there is still room for the “narrow” approach. It

TrinJ 6:2 (Fall 1985) p. 133

remains one moment within a richer total process of interpretation. Now we come back to the question, What is “grammatical-historical exegesis”? This key phrase has a certain looseness. Sometimes it has merely designated the “narrow” approach. Sometimes, I suspect, people have used it loosely, when they had in mind all of the “middle-of-the-road” approach. But many times, at least, the “narrow” approach is what people have in mind. Even if we want to use the key phrase “grammatical-historical” more broadly, we must somehow retain room for the narrow approach, or else we have quite thoroughly evaporated the meaning of “grammatical-historical.”

In short, the narrow approach is not all there is. But it is a part, even an important part. Moreover, I believe it has a special role in dispensational interpretation. Dispensationalist interpreters have laid particularly heavy stress on the importance of grammatical-historical interpretation of OT prophecy. They have complained that nondispensationalists illicitly read the NT back into the OT. In our terminology, their complaint says that nondispensationalists have confused or mixed canonical interpretation with “narrow” interpretation. But when it comes to Dan 9:25, things change. Dispensationalists have almost uniformly adopted the date of 444 B.C., and have advocated the date partly on the basis of New Testament information about the time of the death of Christ. Hence, there is a need to assess carefully what results when “narrow” interpretation is applied to Dan 9:25.

3. “Narrow” interpretation of Daniel 9

Let us then ask what information Daniel 9 gives us, when interpreted “narrowly” against its original historical context. It is important first that we know when Daniel was written, and this is itself debated. I cannot hope to deal with this debate in full, so I will simply assume (following conservative interpreters) that Daniel was written in the sixth century B.C.

Under this assumption, Meredith G. Kline has given a strong argument that Daniel 9:25 points clearly to the decree of Cyrus.⁴ That argument, it turns out, is further strengthened when we pay careful attention to the limits of “narrow” interpretation.

First of all, as Kline points out, several factors in the immediate context of Daniel 9 lead one to expect that this decree will be issued not long after Daniel’s prayer.

The text of Daniel 9 is a single unit with three main parts. Dan 9:1–2 introduces the setting for Daniel’s prayer. Dan 9:3–19 gives the contents of the prayer. The final section, 9:20–27, includes as its main part Gabriel’s announcement to Daniel about the 70 weeks.⁵

Dan 9:1–2 indicates that Daniel’s prayer takes place “in the first year of Darius son of Xerxes.” This Darius is either Cyrus the Great himself, called by a different name,⁶ or a contemporary subordinate of Cyrus. In any case, the year in question is the first year of Cyrus’s reign, 538 B.C. Now Jeremiah had prophesied that after 70 years Babylon would be punished by conquest (Jer 25:12) and the nation of Israel would be restored (Jer 25:11; 29:10). The

TrinJ 6:2 (Fall 1985) p. 134

70 years were evidently to be reckoned beginning from 606 B.C., the first deportation to Babylon, when Daniel and his friends went into exile. Daniel saw that Jeremiah’s 70 years were nearly at an end. This motivated the prayer of Dan 9:3–19, a prayer for restoration. Moreover,Jeremiah had also prophesied the fall of Babylon at the end of the 70 years (Jer 25:12). Babylon fell to Cyrus in 539–538. The fall of Babylon also indicated to Daniel the nearness of the promised restoration.

Gabriel’s answer in Dan 9:20–27 is to be understood as an answer to Daniel’s prayer. This follows from the following considerations. (1) Nothing in the context tells the readers that Gabriel’s message is anything other than an answer. (2) The urgency of Daniel’s prayer and the nearing of the time prophesied by Jeremiah make us look for some answer, and no other candidate for an answer occurs in the context. (3) Gabriel explicitly mentions Daniel’s prayer as the occasion for his being sent (Dan 9:22–23). The NIV actually translates, “an answer was given” (9:23), though the corresponding Hebrew word דָּבָר need not be as specific as the English word “answer.”

If, now, Dan 9:24–27 is the substance of the answer, the reader expects that it will touch on the restoration about which Daniel prayed, which is also the restoration about which Jeremiah prophesied. Jeremiah prophesied about a restoration in approximately 537. Gabriel can be expected to give promises whose fulfillment would commence at the same time. In particular, the beginning point of the 70 weeks in Dan 9:25 can be expected around 537.

This expectation is actually confirmed by the contents associated with the beginning point. “The decree to restore and rebuild Jerusalem” in 9:25 corresponds to Jeremiah’s prophecy “that the desolation of Jerusalem would last seventy years” (Dan 9:2). And it corresponds to Daniel’s prayer. He prays for the end of desolation for Jerusalem: “turn away your anger and your wrath from Jerusalem, your city, you holy hill” (9:16); “see the desolation of the city that bears your name” (9:18); “making my request to the Lord my God for his holy hill” (9:20).

The logical conclusion from this language is that the beginning point of the 70 weeks basically coincides with the end of Jeremiah’s 70 years. That is, it occurs in 538 B.C. or shortly thereafter. One the other hand, a beginning point in 444 B.C. would not really answer Daniel’s prayer. It would not be quick enough to satisfy Daniel’s urgency. And it would not be related to the basis of Daniel’s prayer in Jeremiah’s prophecy of 70 years.

4. Cyrus’s decree in prophecy and in history

Thus far we have reasoned on the basis of information that would have been publicly available to Daniel and his contemporaries at the time when Daniel uttered his prayer. Now we need to ask, “Was there any decree issued shortly after this time, which might correspond to Daniel’s expectations?” The answer is yes.

In the first year of Cyrus kind of Persia (probably 538 B.C.), Cyrus issued a decree about the rebuilding of the temple in Jerusalem:

The Lord, the God of heaven, has given me all the kingdoms of the earth and he has appointed me to build a temple for him at Jerusalem in Judah. Anyone of his people among you—may his God be with him, and let him go up to

TrinJ 6:2 (Fall 1985) p. 135

Jerusalem in Judah and build the temple of the LORD, the God of Israel, the God who is in Jerusalem. And the people of any place where survivors may now be living are to provide him with silver and gold, with goods and livestock, and with freewill offerings for the temple of God in Jerusalem (Ezra 1:2–4; cf. 2 Chr 36:23).

2 Chr 36:21–22 indicates that Cyrus’s decree began the restoration prophesied by Jeremiah. Thus it corresponds explicitly to Daniel’s concern m praying on the basis of Jeremiah (Dan 9:2).

Moreover, the restoration set in motion by Cyrus agrees with the description in Dan 9:25. The decree given in Daniel is “to restore and rebuild Jerusalem” (Dan 9:25). Isaiah had prophesied that this was what Cyrus would do:

[God] says of Cyrus, “He is my shepherd and will accomplish all that I please; he will say of Jerusalem, ‘Let it be rebuilt,’ and of the temple, ‘Let its foundations be laid’“ (Isa 44:28).

I will raise up Cyrus⁷ in my righteousness: I will make all his ways straight. He will rebuild my city and set my exiles free, but not for a price or reward, says the LORD Almighty (Isa 45:13).

Now we have accumulated a considerable amount of material pointing to the conclusion that the beginning point in Dan 9:25 was in Cyrus’s decree. The crucial question is, what does narrow interpretation of Dan 9:25 do with this material? Narrow interpretation asks what can legitimately be inferred from this text in its original context. What information existed in the environment at the time? By the time that Daniel’s material was put in written form, Cyrus’s decree had probably already been issued. If not, it was to be issued within a few months. The original addressees therefore potentially had (1) Cyrus’s public decree; (2) the prophecy about Cyrus in Isa 44:28; (3) the prophecies of Jeremiah indicating restoration after 70 years; (4) the literary context in Daniel 9, indicating that Gabriel’s message was an answer to Daniel’s prayer and therefore to Jeremiah’s prophecy.

On the other hand, the original addressees knew nothing yet about Nehemiah and Artaxerxes’s letter of Neh 2:8. Neither Nehemiah nor Artaxerxes had even been born! Hence narrow interpretation must effectively exclude reckoning with Nehemiah’s time. All the evidence actually available at the time would point interpreters to the conclusion that Dan 9:25 refers to Cyrus’s decree. No evidence then available leads to Nehemiah’s time (444 B.C.).

Hence, the alternative is quite clearly posed: either follow narrow interpretation and choose the date 538 B.C.; or abandon narrow interpretation in the light of needs to fit Dan 9:25–27 into the supposed constraints imposed by later events and by a prophetic system.

But if the conclusion is so clear, how is it that some ever came to advocate the date of 444 B.C.? Well, there is an argument in favor of this date. We will examine it in a moment. But the argument has sprung up, I believe, in a context where the primary mode of interpretation was not really narrow interpretation. Rather, people used the full knowledge of the dates of various historical events subsequent to Daniel’s time.

TrinJ 6:2 (Fall 1985) p. 136

5. The Question of Rebuilding Jerusalem

The main argument for the 444 B.C. date starts from a supposed distinction between the restoration of the temple and the restoration of the city of Jerusalem. Cyrus’s decree, it is said, touched only on the restoration of the temple. But the beginning point in Dan 9:25 has to do with the restoration of the city. Hence we must look for the issuing of a decree having to do with the city, and specifically including the building of fortifications (cf.חָרוּץ, “moat” or “trench” in Dan 9:25). The wall building of Nehemiah, commissioned by Artaxerxes (Neh 2:7–8), matches this requirement.

In answer to this argument, we must first take up several points related to the scope of Cyrus’s decree.

(a) In narrow interpretation, we must bear in mind that the Israelites lived in an atmosphere where the restoration of the temple, the restoration of the city of Jerusalem, and the restoration of the land itself were closely bound up together. The city represented the heart-beat and security of the land around; the temple represented the heart-beat and security of the city (Jer 7:4). Jeremiah prophesied desolation for the land, for the city of Jerusalem, and for the temple. In particular, Jeremiah’s prophecy concerning 70 years of desolation speaks explicitly of restoration of the people to the land (Jer 29:10, 14), but is naturally interpreted to imply restoration of the city (Dan 9:2, 16, 18) and of the temple (Dan 9:17).

(b) Cyrus’s decree, as recorded in Ezra 1:2–4 and 2 Chr 36:23, focuses on the rebuilding of the temple. But we do not have the complete text of the decree here. Ezra 6:3–5, an alternate report of the decree, contains some details not recorded in Ezra 1:2–4. Perhaps still more details exist that have not been included in either summary. Josephus the Jewish historian reports the contents of Cyrus’s letter to the satraps of Syria as follows:

King Cyrus to Sisines and Sarabasanes, greeting. To those among the Jews dwelling in my country, who so wished, I have given permission to return to their native land and to rebuild the city and build the temple of God of Jerusalem on the same spot on which it formerly stood. (Jewish Antiquities 11.12 [ 11.1.3], Loeb edition; emphasis is mine. Cf. also Jewish Antiquities 11.6 [11.1.2]).

Thus Josephus maintains that Cyrus’s decree included reference to the city, not merely the temple. Of course,Josephus may be conflating the decrees of Cyrus and Artaxerxes, but this would be just a supposition.

Besides this, consider that the whole of Ezra 1–8 has a heavy emphasis on the building of the temple and the restoration of worship. We cannot be sure whether details concerning the larger concerns of the city have been omitted. But let us even grant, for the sake of argument, that Josephus was wrong and that Cyrus’s decree contained no explicit word about the restoration of the city. Such a restoration of the city, at least on a humble level, would nevertheless be presupposed as an accompaniment to the restoration of the temple. For one thing, there would have to be workers there in the city to engage in the restoration work on the temple. And the temple would make little sense without a body of priests to serve in it. Some priests would have to be settled in Jerusalem.

(c) Dan 9:2 says explicitly that Daniel understood that “the desolation of Jerusalem [the city] would last seventy years.” In narrow interpretation,

TrinJ 6:2 (Fall 1985) p. 137

there is nothing to lead one to think that Daniel understood incorrectly. Hence restoration of the city, not merely of the temple, was begun after the 70 years. To say that the restoration of the city had to wait until Nehemiah’s time is a denial of the validity of Jeremiah’s prophecy.

(d) Isa 44:28 predicts that Cyrus will speak a word respecting both the temple (“‘Let its foundations be laid,’“) and the city (“he will say of Jerusalem, ‘Let it be rebuilt’“). Now the word בָּנָה (“build”), applied to the building of Jerusalem occurs in Isa 44:28 placed in Cyrus’s mouth.⁸ The same word occurs in Dan 9:25 as the word in the decree. Hence, the claim that Cyrus was concerned only with the temple contradicts Isa 44:28 and overlooks the obvious parallel between Dan 9:25 and Isa 44:28.

(e) Once again, the opponents must be pressed to remember that the information available to us in Nehemiah was not available in Daniel’s time. By contrast, Dan 9:2 and Isa 44:28 were available. Dan 9:2 and Isa 44:28 were bound to be seen as linked. The information in Nehemiah was not available to promote an interpretation in the other direction. It is true that, after a number of years had passed, people would have been able to observe that the gates and walls of Jerusalem had not, in fact, been restored. But it would have taken a number of years before the difference between the small beginnings and what was potentially possible would have been driven home. In the meantime, people were already in a position on the basis of Dan 9:2 and Isa 44:28 to draw their conclusions about the terminus a quo, the beginning of the 70 weeks. Narrow interpretation demands that we stay with the meaning in terms of that context, the context of Cyrus’s decree, not a context a hundred years later. Information that we obtain later may supplement our understanding of a text, but should not be used to undermine the previous understanding.

6. Was Jerusalem reinhabited before 444 B.C.?

It should also be noted that there is evidence that Jerusalem was reinhabited as a consequence of Cyrus’s decree. Neh 3:20, 21, 23, 24, 25, 28, 29, and 7:3 indicated that there were houses in Jerusalem before the start of Nehemiah’s restoration project. Hence it appears that the restoration of the temple and of the city did take place roughly together. The old walls and the gates evidently remained in ruins until Nehemiah’s time (Neh 1:3; 2:8, 13, 17).⁹ But saying that the city had no walls is not the same as saying that it had not begun to be restored.

Over against this, John F. Walvoord appears to take an extreme position,

TrinJ 6:2 (Fall 1985) p. 138

claiming that before Nehemiah’s time “the children of Israel had built houses, but apparently they were not in Jerusalem.”¹⁰ In justification of this he cites Neh 2:12–15, 17, and Neh 11:1.

But Walvoord’s texts do not show what he thinks that they show. Let us begin with Nehemiah 2. Nehemiah 1–2 is almost entirely oriented to the question of the gates and walls of Jerusalem. The whole sequence of events begins with a report to Nehemiah that “the wall of Jerusalem is broken down, and its gates have been burned with fire” (1:3). If, in actual fact, Jerusalem had been completely uninhabited and without houses, the lack of houses would have been an added factor in the distress. But no mention is made of such a circumstance. After this point Nehemiah’s plans unfold in connection with his desire to rebuild the walls and the gates. Nothing is said right away about repopulating Jerusalem and building houses for the people. Nehemiah’s project is expressed in general terms in 2:5 as a project of “rebuilding,” but in the light of 1:3 and 2:6–9 it is evident that the rebuilding need not encompass more than city walls and city gates (and his own residence, Neh 2:8).

Then the text reports that Nehemiah “went to Jerusalem” and stayed there three days (2:11). Evidently he regularly spent the night inside the boundaries of the city, since his night inspection began with him going out (2:13) and ended with him coming in (2:15). Reading between the lines, we can say that there was at least enough cleared space within the city for Nehemiah and those with him to lodge there. Moreover, one gate was open for traffic (2:13). But, as might be expected in a partially rebuilt city, not yet inhabited to the full extent that it was before, another gate (or way to the gate?) was still blocked up (2:14). Walvoord reads the negative features of Neh 2:12–15 as if they were a comprehensive picture of the whole city. he says “[it] pictures the city in utter ruins … the streets [all of them?] so full of debris that his beast which carried him could not get through.”¹¹ But this is an over reading. Nehemiah 2 mentions only one place blocked up with debris (2:14). And it says nothing to imply that there were no houses built in Jerusalem.

There is evidence in Nehemiah 3 that Jerusalem was already inhabited. Neh 3:20, 21, 23, 24, 28, 29 mention individual houses opposite which repair work was done. Neh 3:25 mentions “the upper palace.” Moreover, the entire rebuilding project was completed within a few months of Nehemiah’s arrival (Neh 6:15), despite opposition. The fast work does not seem to indicate that this involves a complete process of repopulation and rebuilding of individual dwellings.¹²

Neh 7:4, describing the situation after the completion of the wall-building, says, “Now the city was large and spacious, but there were few people in it, and the houses had not yet been rebuilt.” Does this mean that there wereno houses? The previous verse, Neh 7:3, has this to say: “Also appoint residents of Jerusalem as guards, some at their posts and some near their own houses.” It mentions both houses and residents. Hence Neh 7:4 means only that large spaces of the city grounds were still open and not

TrinJ 6:2 (Fall 1985) p. 139

rebuilt. There were only houses for the “few” people already dwelling in Jerusalem. Hence a repopulation project was subsequently undertaken (Neh 11:1).

Other texts show that Jerusalem was inhabited even before Nehemiah’s first visit. Let us begin with the restoration in the time of Cyrus. Where did the people settle who, under Zerubbabel, were engaged in building the temple (Ezra 3)? Where would the priests have lived who ministered in the temple? Some indeed lived in surrounding towns (Neh 11:3 20, 12:27–29), just as they had in pre-exilic times. But is it not artificial to imagine thatno one lived in Jerusalem? 1 Esdras 5:45 reports that in Zerubbabel’s time the priests, Levites, and some of the people settled in Jerusalem as well as in the surrounding country. 1 Esdras 5:45 may even represent the original text of Ezra 2:70 (cf. for example the judgment of the RSV translators). Even if it does not, it may represent an accurate historical datum. The burden of proof falls on those who would show the opposite.

By the time of Darius (about 520 B.C.), there were certainly some living in Jerusalem. Ezra 5:1 says that Haggai and Zechariah “prophesied to the Jews in Judah and Jerusalem.” That alone should be weighty enough. Hag 1:4, 9 indicates that at that time private homes had multiplied. Haggai does not say explicitly that the houses were in the city of Jerusalem. But he contrasts the state of those houses with the state of God’s house. It is most natural to assume that he is speaking in the first instance to the inhabitants of Jerusalem. Ezra 6:9 mentions “the priests in Jerusalem.” Ezra 4:6, bringing us to the time of Ahasuerus = Xerxes (486 B.C.), mentions “an accusation against the people of Judah and Jerusalem” (literally, against the inhabitants of Judah and Jerusalem).

Ezra 7–10, describing the situation a few years before Nehemiah’s time, confirms our suppositions. It certainly sounds as if, in Ezra 7:7, 14, 24, 9:9, 10:6, and 10:7, Ezra and his companions came to a Jerusalem which was not completely empty. Ezra 10:6, in particular, says that “Ezra withdrew from before the house of God and went to the room of Jehohanan son of Eliashib.” Here was a room in Jerusalem. It was not merely one of the priest’s rooms within the temple, since Ezra “withdrew from before the house of God” in order to go there. Ezra 10:7 says that “a proclamation was then issued throughout Judea and Jerusalem …,” which seems mysterious if no one was currently living in Jerusalem,

Walvoord¹³ has one more text on this point, namely Neh 11:1. But this text also does not show all that he thinks it shows. He seems to think that it demonstrates that Jerusalem was uninhabited before. All that it explicitly shows is that the population of the city was built up to a larger figure (as Neh 7:34 also implies). Neh 11:1 has two parts, one about the leaders and one about the rest of the people. The KJV connects these parts using an “also,” thereby seeming to imply that the leaders and the tenth of the people all moved to Jerusalem for the first time. But there is no “also” in the Hebrew. One could translate,”the leaders of the people dwelt in Jerusalem. But the rest of the people cast lots to bring one out of every ten to live ha Jerusalem … “That does not necessarily imply that no one at all (in particular, no one of the leaders) was dwelling in Jerusalem up until that

TrinJ 6:2 (Fall 1985) p. 140

time. At an earlier time than Neh 11:1, Neh 7:3 mentions explicitly “residents of Jerusalem” and “houses.” Neh 7:4 says explicitly that there were “few people in it [Jerusalem],” implying that there were some.

One additional little piece of information contributes to the picture. Neh 12:25–26 provides the information that Mathaniah, Bakbukiah, and a number of other Levites served as gatekeepers. “They served in the days of Joiakim son of Jeshua, the son of Jozadak, and in the days of Nehemiah the governor and of Ezra the priest and scribe.” What time period does this represent? Joiakim, according to the genealogy in Neh 12:10–11, was the next high priest in succession after Jeshua = Joshua the son of Jozadak, the contemporary of Zerubbabel (Neh 12:1). Jeshua lived at the time of the restoration under Cyrus in 536 B.C. (Ezra 3:1–2). Neh 12:26 is therefore saying that Mathaniah and others were functioning as gate keepers considerably before Nehemiah’s time, as well as during Nehemiah’s time (444 B.C. and after). Now the job of gatekeeper involved guarding the storerooms at the gates of the temple (Neh 12:25). It is hard to imagine that that job was done by a group of men, none of whom lived in Jerusalem until Nehemiah’s time.

On the basis, then, of a considerable amount of direct and circumstantial evidence, we conclude that Jerusalem was (partially) inhabited before Nehemiah’s time. In Nehemiah’s time the people accomplished the building of the walls and city gates (Nehemiah 3–6), and further repopulated the city (Nehemiah 11). If one does not admit this, one comes into conflict with Dan 9:2 and Isa 44:28 as well as with some of the verses in Ezra and Nehemiah that we have cited.

7. The argument for finding wall-building in Dan 9:25

Another argument bases itself upon the word חָרוּץ in Dan 9:25b. The KJV translates this word as “wall”; NASB has “moat”; NIV has “trench.” The argument therefore is to the effect that the decree in view in Dan 9:25a is a decree specifically including within its scope the building of fortifications, walls and gates. Hence one has to wait for the decree of Artaxerxes (444 B.C.) to find the fulfillment.

How do we evaluate this argument? Like some of the other arguments for the 444 B.C. date, this argument creates serious tensions with the unity of Daniel 9. We have already noted (section 3) that the nature of Daniel’s prayer leads us to expect an answer in the near future (corresponding to Jeremiah’s 70 years), an answer specifically related to the concerns of the prayer. Unfortunately, the rebuilding of the fortifications is neither an event of the near future nor an event that directly answers Daniel’s prayer.

On the other hand, unlike the other arguments above, this argument does at least avoid the obvious tensions with Dan 9:2 and Isa 44:28. It allows that Cyrus’s decree can inaugurate a “rebuilding” of Jerusalem in a general sense, as long as this does not include city fortifications. For this reason, the argument deserves serious attention.

The major difficulty in the argument is that it overlooks the functional difference between the decree of 9:25a and the historical results described in 9:25b. Dan 9:25a (“… from the going forth of the commandment to

TrinJ 6:2 (Fall 1985) p. 141

restore and to build Jerusalem … “) describes the nature of the verbal act marking the beginning point of the 70 weeks; Dan 9:25b (“the street shall be built again, and the moat … “) describes the nature of nonverbal historical results. The historical results may partly be a result of the decree; but they may exceed the bounds of what the decree says (or, hypothetically, if the decree were ineffective, they might fall short of what the decree specified). Moreover, the text in 9:25 says nothing about how long it will take for the rebuilding to be completed.¹⁴ It does not say whether fortifications will be the result of a second or third stage in a long process of rebuilding. The dating proceeds from the original or initial issuing of the decree. That also leaves open the question of whether there might be later decrees or letters dealing with specific aspects of the rebuilding.

In addition to this major difficulty, there is one more point that is at least a minor irritant to the position we are examining. It involves the question of the correct translation of the word חָרוּץ. The KJV (the familiar version for popular dispensationalism) renders the word by “wall,” producing a striking English verbal parallel with Nehemiah 1–3. But this is not a correct translation.¹⁵ The Hebrew word חָרוּץ occurs only here in the Hebrew Bible (ignoring homonyms). Hence there is some uncertainty about its meaning. But, on the basis of cognate languages, it appears that the most probable meaning is “moat” or “trench.” A moat is certainly one part of city fortifications. Walls are another part. But the two are not the same. The book of Nehemiah actually gives no information at all either about a decree concerning a moat, or labors to construct one. Neither does it give any information about a decree concerning a plaza or plazas (“plaza” being the other key word in Dan 9:25b). Perhaps work on “plaza and moat” took place somewhat later than Nehemiah’s time. Theoretically, such work could also have taken place earlier than Nehemiah’s time! In fact Ezra 10:9 indicates that a plaza existed in front of the temple even before Nehemiah’s visit. (Ezra 10:9 uses the same word for “plaza” as Dan 9:25b.) Beyond that, we simply do not have definite information. Moreover, it is at least possible to understand the word חָרוּץ as denoting trenches for water-supply or drainage, rather than a moat.¹⁶ Thus the whole construction is fraught with difficulty.

TrinJ 6:2 (Fall 1985) p. 142

8. Does Dan 9:25 suggest a whole series of decrees?

One final argument relies on the absence of the definite article accompanying the word “decree” (דָּבָר) in Dan 9:25. One might argue that this means only “a decree,” “some decree” (possibly out of several decrees of this kind). Thus, the fact that Cyrus’s decree matches the description does not mean that it is necessarily the one in question. There may be several more that fit the case. We can know which decree it is only from a much later standpoint in history and revelation.

But this argument misunderstands the range of cases where the definite article is regularly omitted. Gesenius’s grammar (section 125i) notes: ∙… the article is always omitted when a person or thing is to be represented as indefinite (or indefinable) or as yet unknown; …” Since the decree is still unknown to human beings at the time when it is first mentioned to Daniel in Dan 9:25, it appropriately has no definite article. Moreover, the omission of the article is regular,when דָּבָר is followed by an infinitive construct clause providing the contents of a message or condition: 2 Sam 15:28; 2 Chr 30:5; Ezra 8:17; Deut 15:9; Isa 30:21; 2 Sam 3:13; 1 Kgs 2:30; 12:16; Ezek 9:11. Thus the absence of the article in Dan 9:25 needs no special explanation. It does not make the decree one of a possible series, any more than is the case with the decree in 2 Chr 30:5. Moreover in the particular case of Dan 9:25 the context of answering Daniel’s prayer indicates that one particular decree, not a whole series of possibilities, is in view. The passage creates no tendency to look for a second decree after it is found that Cyrus’s decree matched.

9. The “weeks” of Daniel 9

There is still one serious obstacle, however, to the date of 538 B.C. Is this date compatible with the actual content of the prophetic predictions in Dan 9:24–27? If the events did not turn out as predicted, then what?

I believe that the prophecies of Daniel do square with later history. But to see this, we must challenge the assumption that the “weeks” in Daniel 9 are weeks of ordinary chronological years.

To begin with, nowhere does the text of Daniel actually say that the “weeks” of Dan 9:24–27 are weeks of ordinary chronological years.¹⁷ In defense of the idea that these weeks are each seven ordinary years, appeal is sometimes made to the mention of years in Dan 9:2, in the context of Jeremiah’s prophecy of seventy years.¹⁸ How might this be relevant? Well, Jeremiah’s 70 years are all sabbatical years, years when the land rests (2 Chr 36:21, based on Lev 25:1–7). Since each year is in effect a seventh year, a total of 490 years or ten jubilees is involved. Hence there is indeed a close relation between Jeremiah’s 70 years and the 70 weeks = 490 periods of Dan 9:24–27.¹⁹ It is claimed, then, that the period of 70 weeks must be a period of ordinary years, just as Jeremiah’s 70 years was a period of ordinary years.

TrinJ 6:2 (Fall 1985) p. 143

But the relationship between Jeremiah’s 70 years and Gabriel’s 70 weeks may be more complex than that: it may be a relationship of analogy rather than pure identity. Moreover, there is a crucial difference between Jeremiah and Daniel 9. Jeremiah’s prophecies (Jer 25:11–12, 29:10), as well as the reference to them in Dan 9:2, exhibit the normal method in the OT of talking about ordinary chronological years, namely, by the inclusion of the Hebrew word שָׁנָה “year.” Daniel 9:24-27, by contrast, does not use the word “year.” When the word “week” (heptad) occurs without the qualifying word “year,” it commonly designates weeks of days (Gen 29:27–28; Exod 34:22; Deut 16:9; 2 Chr 8:13; Jer 5:24).²⁰ Hence there is actually grammatical-historical reason to think that these may not be weeks of ordinary years. If the writer had wanted to indicate years, he would normally have included the word “year,” as in Dan 9:2. Those who argue that weeks of ordinary years are involved here can really give no reason why the word “year” should have been omitted. The omission is contrary to what appears to be the regular grammatical practice in OT Hebrew.

Hence, there is no firm grammatical-historical reason for saying that the weeks are weeks of years. But the weeks need not be weeks of days either. At the heart of the temporal language in Dan 9:24–27 is a sabbatical and jubilee pattern.²¹ This pattern as an OT symbolical pattern is not confined to either years or days. In fact, there is also a kind of “sabbatical” month, the seventh month (Lev 23:23–43). In principle, the word “week, heptad” might also be used to designate a period of still another length, if that period were viewed as related to the sabbatical pattern.

This is as far as we can of if we permit only evidence from the Hebrew OT canon. However, other ancient materials written in Hebrew are relevant, if they show that there is a regular or customary usage of the word “heptad.”

One such evidence comes from the pseudepigraphical book 1 Enoch, written several centuries after Daniel. In 1 Enoch 93:3–10; 91:12–17 the whole history of the world is divided up into 10 “weeks”; and the weeks have varying length when measured in terms of ordinary years. The parallel between 1 Enoch and Daniel 9 is all the more significant because (a) 1 Enoch represents roughly the same genre of literature as Daniel 9 (apocalyptic); (b) 1 Enoch and Dan 9:24–27 use a numbered pattern of weeks for the similar purpose, to outline the course of world history up to the renewal of heaven and earth (1 Enoch 91:16); (c) both have the same verbal construction with “week” not modified by an accompanying specification of “years” or “days”; (d) an Aramaic fragment of I Enoch from Qumran attests that the underlying Aramaic is indeed שׁבוע, the equivalent of the Hebrew of Dan 9:24–27 (see 4QEn^g 1 iv 15, 19, 25).²²

If we move to a time several centuries after 1 Enoch, we can obtain some evidence in the opposite direction. In the Talmud “week(s)” alone can be used to designate the seven-year cycle (cf. b.Ned. 60a, 61a).²³ But this usage occurs at a time when the rabbis had developed a large repertory of abbreviated and technical usages for discussing matters related to the OT.

TrinJ 6:2 (Fall 1985) p. 144

Moreover, the pseudepigraphical Book of Jubilees shows that a calculating mentality had developed in certain circles. Rabbis who did interpret the “weeks” of Dan 9:24–27 as weeks of years (cf. b. Yoma 54aa) could easily have used Daniel itself as a starting point for the technical usage of “weeks.” Hence, it is questionable to assume without further evidence that such a usage was established in Daniel’s time.

The Book of Jubilees (probably some 400 years after Daniel) shows a kind of intermediate usage. The word “weeks” frequently occurs with the meaning “weeks of years.” But almost always this is in the context of a fixed construction of the form, “in the third week in the second jubilee” (Jub 4:1), “in the thirty-fifth jubilee, in the third week, in the first year thereof” (Jub 11:1), “in the fourth year of this week” (Jub 14:1), and the like. The accompanying words have the effect of tying down the meaning of “weeks.” Moreover, the cycles of seven are used throughout the Book of Jubilees as a detailed chronological system. Hence the over-all context in Jubilees also helps to specify the meaning of “weeks.” In view of the lack of equivalent definiteness in Daniel 9:24–27, it remains an open question whether Daniel 9:24–27 is intended to form a chronological scheme in exactly the same way as the Book of Jubilees.

Some of the Qumran texts (probably some 400 years later than Daniel) show three instances of an intermediate usage. 1QS 10.7 and 10.8 have the word שׁבועיחם, “their weeks,” designating seven-year units. But the modifying pronoun “their” refers to years, so that the expression is actually equivalent to “weeks of years.” CD 16.4 has a similar expression “their weeks.” designating seven-year units. But the modifying pronoun “their” refers to years, so that the expression is actually equivalent to “weeks of years.” CD 16.4 has a similar expression “their weeks.” The modifying “their” refers to “times,” but the accompanying parallel word “jubilees” helps to define these “times” as years.

All of this evidence, pro and con, is considerably later than Daniel. It is not part of the OT. We must exercise caution in applying it to Daniel’s time. but if we allow ourselves to appeal to the late evidence for the use of “weeks” to designate weeks of years, we must equally allow ourselves to appeal to the (somewhat earlier) late evidence for the use of “weeks” for indefinite periods viewed as forming a sabbatical pattern (1 Enoch). The evidence tends, in general, at least to confirm the notion that the word “weeks,” in itself, is simply not that precise. A writer’s use of the word does not commit him to rigid agreement with other writers.

We can only deal with possibilities at this point. 1 Enoch is evidence for the possibility of using “weeks” of varying chronological length. The original readers of Daniel could not have excluded this possibility. They did not have the right simply to assume that weeks of ordinary years were involved. I think that from the standpoint of “narrow” interpretation, there is reason to think that the “weeks” of Dan 9:24–27 were probably not simply weeks of ordinary years. (Why else is the crucial word “years” omitted?) However, to establish even the possibility of this option is sufficient for my purposes.

It is legitimate to inquire why the period of 70 weeks is broken up into sub periods of seemingly definite length (7, 62, 1).”Surely,” someone will say,

TrinJ 6:2 (Fall 1985) p. 145

“this must mean 7 x 7: 49 years, then 62 x 7 — 434 years, then 1 x 7 =7 years.” But Kline’s interpretation of Daniel 9 shows that there is another possible motivation for the sub periods.²⁴ The first period of 7 weeks constitutes a single jubilee; thus it may suitably signify the time leading to a preliminary, provisional form of restoration. The entire cycle is 10 jubilees, signifying the time leading to the final restoration. Now the end of the 69th week is just short of this final restoration; it thus suggests a period of inaugurated eschatology. The number 62 is a figure of no interest in itself, but simply the result of 70 – 1 – 7.

10. The impressive precision of the 444 B.C. date

Finally, we must say something about the remarkable precision established in calculations by Sir Robert Anderson and Harold Hoehner.²⁵ Anderson claims to show that the time from Artaxerxes’ decree to Jesus’ Palm Sunday entrance into Jerusalem comes out to exactly 70 weeks of years, even down to the day. This is certainly impressive, at first sight. But to make this calculation work, Anderson and Hoehner after him must use “years” that are all 360 days long.

If one keeps firmly in mind the goals of narrow interpretation, the untenability of years of 360 days each becomes apparent. Put yourself in the shoes of the OT Israelite. Think of how the number of years of reign of Israelite kings was reckoned. Think of the number of years in genealogies like Gen 11:10–26. Think of how the Israelites were intended to count the years to the next jubilee. Does anyone seriously want to contend that figures of this type were intended to be understood in terms of 360-day years instead of solar years?

It is true that Israelites thought of any single year as 360 days long. It had 12 months of (roughly) 30 days each. But even this was not an exact figure. The beginning of each month was at new moon, and the lunar month is only 29.5 days. Hence an actual “year” of 12 lunar months would run, on the average, 354 days. In order to keep the calendar from drifting, Israelites periodically introduced an extra intercalary month between two yearly cycles. The question then is, “Did Israelites include these intercalary months when they calculated long periods of years?” Of course they did.

The case of reckoning to the next sabbatical year is particularly relevant. A sabbatical year occurred every seventh year (Lev 25:1–7, cf. Deuteronomy 15). If the interpreter works with (rough) solar years, then the sabbatical year is simply the 7th solar year, during which there is to be no agricultural cycle of sowing and reaping (Lev 25:4–5). On the other hand, suppose that the sabbatical years are to be reckoning in terms of “years” of 360 days each. The six years of normal agricultural work (Lev 25:3) make up 6 x 360 = 2160 days, about a month short of the approximately 2191.5 days involved in 6 solar years. Hence, if this schedule is followed rigorously, the time during

TrinJ 6:2 (Fall 1985) p. 146

the agricultural year when the rest is to begin must shift around gradually, from one sabbatical year to another. Clearly this is not only a wildly fanciful interpretation of Leviticus 25, but it is agriculturally and socially unworkable. The sabbatical year was calculated in terms of (rough) solar years. The intercalary months, used by the Jews to keep the calendar from drifting, were to be included rather than excluded. Now what about the jubilee year? It must also be calculated in terms of solar years, since it is expressly said that the people are to “count off seven sabbaths of years” (Lev 25:8). The jubilee calculation is thereby tied in with the calculation of sabbatical years.

Now Daniel 9, according to the dispensational approach and that of Anderson, contains 70 sabbatical years or 10 jubilee periods of years. If the original addressees did decide that Daniel 9 was talking about ordinary years, there was every reason for them to think that this jubilee pattern followed the pattern of Leviticus: it was a pattern of solar years.

Anderson, McClain, Hoehner,²⁶ and others postulate a special “prophetic year” of 360 days partly by appeal to Rev 12:6, together with a series of correlations between Revelation and time units in Daniel. But (a) the appeal to Rev 12:6 goes outside the bounds of “narrow” interpretation. (b) The figure of 1260 days may well be only a round number, a rough figure. Note that the 1260 days are also described as 42 months (Rev 11:2–3). The particular figure of 1260 is obtained by multiplying the figure of 42 by the number of days in a month, 30 days. By using an exact multiple, the book of Revelation makes it more obvious that the two time periods are connected. But in actuality lunar months are only 29.5 days, so that the total figure is only approximate. For a short period like this, it would be confusing to introduce a reckoning bases on a intercalary month. Similarly, an intercalary month and short months of 29 days are not explicitly reckoned with in the passage on Noah’s flood. That is why one finds a year of 360 days when one makes calculations based on Genesis 7–9. Only when a longer sequence of years is in view, as in the case of Daniel 9, are intercalary months to be expected as a natural part of the reckoning.

Anderson²⁷ appeals to two other sorts of calculation: (a) there are 70 years of 360 days each from the siege of Jerusalem in 587 B.C. (Ezek 24:1–2) to the laying of the foundation of the temple in 520 B.C. (Hag 2:18); (b) there are 490 years of 360 days each from the dedication of the temple by Solomon (1 Kgs 6:38; 8:2) to the foundation-laying in 520 B.C. These calculations would have theoretically been possible as soon as the events of 520 B.C. had occurred; hence they are not entirely outside the scope of narrow interpretation, particularly if the publication of the Book of Daniel was delayed until nearly 520.

But several factors count against the idea that either of these calculations could have any weight in the interpretation of Daniel 9 by the original addressees of Daniel.

(1) Both calculations depend on Anderson’s²⁸ theory that Jeremiah made

TrinJ 6:2 (Fall 1985) p. 147

prophecies about two distinct periods of 70 years each. The one period was the period of the “Servitude to Babylon” mentioned in Jer 29:10, during which the people of Israel were subjected to Babylonian rule. This dated from the time of the first deportation in 606 to the time of Cyrus’s decree in 536. The second period was the period of “the Desolations of Jerusalem” mentioned in Jer 25:11–12. This was the time during which the land was idle, from 587 to 520. But such a distinction does not really exist, as points (2) and (3) below will confirm.

(2) The supposed distinction between Jer 25:11–12 and 29:10 is invisible unless one makes microscopic distinctions. Both passages speak of the end of Babylonian dominance at the end of 70 years. Why should narrow interpretation ever discern two different beginning points for the prophecies?

(3) If one does read the prophecy of Jer 25:11–12 expecting microscopic precision, it does not fit Anderson’s case. Jer 25:11–12 says nothing directly about the exact length of time during which the land is desolate. The seventy years is a seventy years during which “these nations will serve the king of Babylon.” The end of the seventy years comes immediately before the time when “I will punish the king of Babylon.” Both the end of service to Babylon and the punishment to Babylon can coincide with no other date than that of the fall of the Babylonian kingdom in 538. Thus Jer 25:11–12 coincides both in dating and in content with Anderson’s “Servitude to Babylon,” not with “the Desolations of Jerusalem” as he alleges.

(4) Anderson’s endpoints for the supposed period of “the Desolations of Jerusalem” are not defined in a consistent way. He takes the beginning point to be the time when the Babylonians laid siege to Jerusalem, which would have “put an end to all agricultural pursuits.”²⁹ On the other hand, the endpoint is the time of laying the foundation of the (second) temple (520). this is mixing apples and oranges. For consistency, one must choose: either date both beginning and end in terms of the presence of agricultural work in the land, or date both beginning and end in terms of the existence of temple or temple worship.

If it comes to the choice, Anderson would doubtless be forced to choose for dating in terms of the land, because he relies on the point made in 2 Chr 36:21. He can still say that blessing to agriculture starts from 520 onward (Hag 2:19). But this is still inconsistent. The beginning point is now dated from the time of actual physical cessation of agricultural activity (587 B.C.), whereas the end point is dated not from the time of actual physical resumption of agricultural activity (about 537 B.C.) but from the time when the activity is blessed (about 520 B.C.). Again one must choose: either date both in terms of physical cessation and resumption or date both in terms of the termination and resumption of blessing.

The more one works with this, the more one sees that Anderson actually had quite a few options for picking dates to form the basis for a mathematically exact calculation. He chose the options that gave him the result he was looking for.

(5) Anderson has almost certainly misinterpreted Hag 2:18–19, in

TrinJ 6:2 (Fall 1985) p. 148

assigning a 520 date to the laying of the temple foundation. The first foundation-laying for the second temple did not take place in 520 but in 536 or thereabouts (Ezra 3:10). Then what about Hag 2:18–19? Keil³⁰ argues that the foundation-laying of 2:18 refers to a time (536) distinct from the “twenty-fourth day of the ninth month.” Baldwin³¹ argues that the building process in question in Hag 2:18 was not literally foundation-laying. In either case. Ezra 3:10 and 6:15–16 represent far more crucial dates in the chronology of restoration. The date given in Hag 2:18 does not enjoy the kind of historical prominence that is to be expected for the terminus of Jeremiah’s seventy years.

(6) In order to get his calculations to bear any weight, Anderson has to assume that the 70 in Jeremiah’s prophecy is an exact number rather than a round number. There is no proof of this (after all, there are round numbers elsewhere in the Bible).

(7) Even if Jeremiah’s number 70 is an exact number, it may be a number that counts as whole years the parts of the year at the beginning and the end of 70 (in a manner parallel to the counting of three days between Jesus’ death and resurrection). That is, if the start of the 70 year period is in (say) the tenth month of 606 B.C., the time from the tenth month to the end of the year may be counted as year 1. Likewise with the final year of the period of 70 years. In that case, Anderson’s 69 years is 70 years, without any necessity of eliminating intercalary months. Hence Anderson’s argument that prophetic dates did eliminate intercalary months remains unsupported.

(8) Anderson’s endpoints for the period of 490 years are again endpoints of two different kinds. Either both endpoints should be the dates of dedication of the temple, or both should be dates of foundation-laying. Interestingly, if one takes both endpoints as the dates of dedication, one gets a period of 1005 to 515, 490 solar years by Anderson’s own admission.³² Anderson “fudges” this figure not only by choosing endpoints of two different kinds, but by reckoning from the year after the dedication of Solomon’s temple to the year before the foundation-laying of the second temple. The movement by a year at each end is certainly a convenience if one wants to make one’s figures come out on the button!

(9) Anderson’s own arguments concerning the 490 years find their starting point in 2 Chr 36:21. 2 Chr 36:21 makes each of Jeremiah’s 70 years a sabbatical year. Hence, Anderson supposes, there must be a period of 490 total years involved, ending in 520 B.C. Of these 70 x 6 = 420 were tillage years, after which came the 70 years of non tillage. Suppose we follow Anderson’s rigoristic interpretation of 2 Chr 36:21. Then the endpoints of the 490 years should not, in fact, be chosen on the basis of what happened to the temple, but what happened agriculturally. The land should rest as many years as it had missed during Israel’s occupation and tillage of the land—not, mind you, as many years as the temple stood. But in fact there is no need to understand 2 Chr 36:21 with this rigorism. Seventy sabbatical years, making

TrinJ 6:2 (Fall 1985) p. 149

up 10 jubilee periods, is a number of completeness. It could correspond to whatever was the complete length of years during which Israel inhabited the land without observing the sabbatical years.

(10) Anderson himself cites two time periods whose length is defined in ordinary solar years.³³ One of these he even insists on as proof of “the absolute accuracy of these Scriptures.”³⁴ He thereby tacitly admits that the interpreters of Daniel’s time would have naturally reckoned the length of these periods in terms of ordinary solar years.

11. Conclusion

In conclusion, “narrow” interpretation of Dan 9:24–27 points decisively to the decree of Cyrus as the beginning point for the 70 weeks. This is inconsistent with the dates of fulfillment only if one insists (contrary to the vagueness of the word “weeks”) that weeks of ordinary literal years must be involved. Moreover, “narrow” interpretation is an aspect of what most people have in mind by “grammatical-historical interpretation.” Hence, it behooves advocates of other views to re-examine either their interpretation of Dan 9:24–27, or else the understanding they have of their commitment to grammatical-historical interpretation.

---