The right doctrine from the wrong texts? – fulfilment in Matthew’s Gospel

juni 4, 2009 at 12:23 am 7 kommentarer

Ever wondered why the New Testament sometimes quotes seemingly obscure texts from the Old Testament and claims that Jesus fulfilled them? Or wondered why the quotations in the New Testament are not giving the exact words of the Old Testament passage which it quotes? In what sense did Jesus in fact fulfil the Old Testament prophecies?

I’ve never really grasped this. I always had a feeling that I got it all wrong, biased by my 2oth century understanding of prophecy. I’ve done a bit of studies in this area which resulted in the paper found below. It tries to understand the quotations in Matthew in light of contemporary Jewish principles of interpretation (i.e. ‘hermeneutics). I’ve not fully grasped it but I think I’m getting closer to understading what in the world is going on in Matthew’s Gospel. Read the paper if you are a nerd – or out of curiosity 😉

I’m writing a final paper which is related to this one. It’ll be a detailed word study on the verb “to fulfil” in the Gospel of Matthew. Let’s see what that brings of new insights …

Matthew’s use of Scripture in Matthew 2:15 and 2:23


Matthew’s formula quotations[1] have been a considerable challenge for modern scholarship leading commentators to remarkably diverse conclusions. This fact is due to the fundamental nature of the subject, for it touches at the heart of Jewish hermeneutics and of the early Christian understanding of prophecy and fulfilment.

This paper will try to answer some of the questions related to the formula quotations in the infancy narrative in Matt 2:15 and 2:23.[2] We will seek to understand the text from Matthew’s point of view in order to draw out the originally intended meaning. This requires that we judge Matthew as an interpreter and narrator on the basis of his own hermeneutical standards. We will see that rather than distorting the text and being arbitrary in his use of Scripture (as the accusations sometimes go), Matthew communicated perfectly clear and reasonably into his own context.

A survey of the literature in the field will help us understand the text within the framework of contemporary scholarship. After that, we will analyze the texts and try to make sense of it, both in the narrative in Matthew and in relation to Scripture. This will help us see the message of the infancy narrative in its traditional context in first century Jewish societies.

Survey of literature

The amount of literature on Matthew’s use of Scripture has grown immensely in the 20th century. This fact alone is evidence of both the difficulty of the problem and that consensus has not been achieved. The views of Matthew’s use of Scripture extend from sensus plenior (fuller meaning)[3] and claims of literal fulfilment[4] on one side, to false citations,[5] unreliable treatment of sources,[6] or seeing the infancy narratives as legendary[7] on the other side. It raises the question on what ground we should evaluate Matthew’s hermeneutic. Judging Matthew on the basis of conventional modern academic criteria is anachronistic and hence methodologically flawed. But valuable attempts of understanding Matthew within his own context have been made. This survey of literature will summarize and evaluate some of the milestones of research in the 20th century.

Before the discovery of the Dead Sea Scrolls (DSS) in 1947, a hypothesis was proposed by Harris who argued that the source behind Matthew’s formula quotations was a Testimony Book containing a collection of proof-texts that showed Jesus’ Messiahship.[8] In the beginning the hypothesis rested on readings of Matthew’s Gospel and evidence of a similar book edited by Cyprian in the early 3rd century. But the discovery of similar texts among the DSS provided evidence parallel to the socio-historical context of Jesus and must therefore be considered significant evidence in favour of the hypothesis. But as Gundry has pointed out, even if true and relevant to the question of redaction, the existence of a Testimony Book does not contribute much to our understanding of Matthew’s use of Scripture.[9]

Stendahl’s considerable contribution in 1954[10] should be seen in light of the ongoing excavations and publications of the DSS. His comparison of the hermeneutics in Matthew and the Habakkuk commentary from Qumran (DSH) led him to conclude that Matthew’s formula quotations are the result of a rabbinic school similar to the hermeneutic community at Qumran within which the DSH emerged.[11] On this basis he labels Matthew’s hermeneutic an early form of midrash pesher, a Jewish method of interpretation in which the fulfilment rather than prophecy was the guiding factor.[12] The effort to understand Matthew contextually is a strong point in Stendahl’s work, but several commentators have criticised him for placing too much emphasis on the similarities between Matthew and the DSH at the expense of the differences.[13] Thus the pesher method builds from and elaborates on the prophetic text, whereas Matthew’s story line takes its point of departure in the historical events as he knew them.[14] Furthermore, there is no need to assume a Matthean rabbinic school comparable to the Qumran community in order to explain the hermeneutical phenomenon. This comes from Stendahl’s form-critical assumptions. However, even if we cannot agree on the assumption that Matthew and the DSH came into being within largely similar social situations, one point of comparison seems to be both valid and most instructive: both Matthew and the Qumran community interpreted ancient texts not from the point of view of the prophetic utterance but from the point of historical fulfilment.

Gundry widened the scope of Stendahl’s study by including what he calls “allusive quotations” in his study. [15] On this basis he concluded that the free style of quoting in the formula quotations is the conventional method rather than a distinct and abnormal feature in the use of Scripture in the Gospels.[16] He believes that this is evidence that the Targumic style was common in the early church.[17] Gundry further argued that Matthew’s hermeneutic principles can be traced back to Jesus himself because all the major themes are found in the sayings of Jesus himself. However, here Gundry confuses theological theme with hermeneutic principle. Just because Matthew and Jesus both present Jesus as the Davidic King, it does not follow that Matthew is justified to use an arbitrary prophetic text to prove that point.[18] Thus Gundry helps us to see that the quotations result from Matthew’s own free rendering but fails to give any specific hermeneutical principles and explain the contextual origin of such principles.

From the overwhelming amount of literature in the last four decades since Gundry, a few contributions, which are representative of a general development towards denying historicity of the infancy narrative, deserve attention. Prabhu[19] and Brown,[20] both Roman Catholics and writing in 1976, argue that Matthew’s main purpose was theological, which leads them to respectively deny and question the historical value of the infancy narrative.[21] Pesch argues in a similar fashion that Matthew’s Gospel (not just the infancy narrative) uses a “messianic exegesis” derived from the “community theology” as experienced in the congregation in Antioch.[22] If the value of Matthew’s Gospel lies in the experiential level of truth among his community rather than the historical event, it virtually makes the question of factuality irrelevant. If indeed the infancy narrative were presented as a-historical,[23] this conclusion would have been a satisfactory solution. However, France has rightly pointed out that an a-historical reading fails to explain basic features of the text. Firstly, why did Matthew build on these particular texts if he had the freedom to invent stories? He could instead have chosen passages more suitable to his purpose. Secondly, if Matthew had the freedom to invent the stories, why did he invent stories with such an obscure relevance for his proof-texts? The answer to both questions can only be that Matthew was guided by historical fact.[24] We therefore conclude that the right question to ask is not about historicity but about historiography.

Text-form and Hermeneutic

One of the problems of understanding Matthew’s use of Scripture is that we do not know which version of Scripture he used when he wrote his Gospel. Did he translate from Hebrew (MT)[25] or Aramaic (Targums), or did he use a Greek translation? And if he used a Greek translation, which version did he use (LXX[26], revised LXX, or translation unknown to us)? Modern versions of Scripture are fairly standardized, but that was not the case in contemporary Judaism. The question of text-form becomes important when we ask how Matthew used Scripture. Are the deviances in Matthew’s quotations due to deliberate alterations or due to the existence of a pre-Matthean text? And if he altered the text, what did he alter it from?[27] And if he altered the texts, should we understand it as manipulation of the text or as a contextually legitimate hermeneutical method?

The problem is that it is very difficult to distinguish text-form from interpretation, especially in a milieu that was trilingual.[28] Translations are always interpretations, and in a tradition with a certain level of hermeneutical freedom such interpretive features would be expected to be particularly evident. We will therefore have to consider both the question of text-form and of hermeneutic in our discussion of Matthew’s formula quotations.

Matthew 2:15 and Hos 11:1

The quotation in Matt 2:15 is the third of five in the infancy narrative of Jesus.[29] Commentators agree that Matthew refers to Hos 11:1, but apart from that no consensus exists. The first problem is the question of the text-form underlying Matthew’s quotation.


Matthew (2:15): ἐξ Αἰγύπτου ἐκάλεσα τὸν υἱόν μου (“from Egypt I have called my son”).

MT (Hos 11:1): מִמִּצְרַ֖יִם קָרָ֥אתִי לִבְנִֽי (“from Egypt I have called my son”).

LXX (Hos 11:1): ἐξ Αἰγύπτου μετεκάλεσα τὰ τέκνα αὐτοῦ (”from Egypt I have summoned his children”).


This comparison shows that Matthew is closest to the MT. The LXX differs from Matthew in the use of the compound verb “to summon” (meta-kaleo; μετακαλέω) rather than Matthew’s “to call” (kaleo; καλέω). Furthermore, the LXX translated the Hebrew “my son” into “his children” while Matthew has retained “my son”. This has led many to conclude that Matthew used his own fresh translation of the Hebrew when he quoted from Hos 11:1.[30] However, if Menken is right that the LXX only became a standardized text-form in the later Christian era, then Matthew’s text could as well be a revision of the Greek LXX translation, which would likely have made the LXX conform to the Hebrew MT.[31] Thus we have two proposed explanations for the similarities between Matthew and the Hebrew text: One sees Matthew’s text as his own fresh translation of the Hebrew MT; the other sees Matthew’s texts as a revision of the LXX, a revision which made the Greek text conform to the original Hebrew text.

Despite the greater similarities between Matthew and the MT, one factor leads us to believe that Matthew used a Greek text-form available to him rather than translated himself. Menken shows that there was an increasing preference for ἀπὸ (from) instead of ἐκ (from) before geographical proper names (such as “Egypt”) from the time of the LXX (ca. 3rd century BC) onwards. Aquila (Greek translation from 2nd century AD) has the exact same rendering as Matthew apart from the preposition ἀπὸ instead of ἐκ before the geographical name ‘Egypt’. The difference between the LXX (ἐκ) and Aquila (ἀπὸ) is due to a development in the use of preposition for the meaning “out of” in a local sense. The question is therefore which preposition Matthew would have preferred had the quotation in Matt 2:15 been his own translation. Menken shows that Matthew has a tendency towards using ἀπὸ when he intends the local meaning “out of”. [32] The use of ἐκ in his quotation in 2:15 is therefore evidence that the translation is not his own but that he was using a Greek translation.

We therefore conclude with Menken that despite the greater similarities between Matthew and the Hebrew MT than between Matthew and the LXX, Matthew most likely used a Greek translation unknown to us. This should not surprise us given the fact that the Greek text-form was not yet standardized at the time when Matthew wrote his Gospel.

Having established that Matthew in this instance probably quoted his source accurately, we will now look at the quotation in its original context in Hos 11:1.

The son in Hos 11:1 clearly refers to Israel as a corporate body. This is further evidenced by the fact that the LXX renders it in the plural (“his children”), thereby interpreting the son to be the whole people of Israel. This is important because it shows that Hos 11:1 was not given a messianic interpretation in the second century BC, for it would have required a translation with a reference to an individual Messiah. Rather than that, Hos 11:1 referred back to the time since the Exodus from Egypt as a reminder of the historical event where God saved his people.

In Hos 11:1, the Hebrew preposition with the geographical name can be understood in two different ways, either temporally or locally.[33] Thus “from Egypt” can be understood as either “from the time of Egypt” (i.e. “since Egypt”) or “from the place of Egypt” (i.e. “out of Egypt”). The context in Hosea speaks in favour of the former temporal sense. The calling is reiterated in Hos 1:2 where it becomes clear that the calling is continuous (NRSV: “The more I called them”) just as Israel’s apostasy was continuous (NRSV: “They kept sacrificing”). The calling cannot be considered a one time event. The emphasis of the preposition in the original context is therefore temporal (i.e. “since the time of Egypt”).[34]

Our conclusion on the verse in its original context is that it was absolutely devoid of any messianic, forward-looking meaning. Furthermore, the preposition should be understood temporally, i.e. “since Egypt”.

When we turn to Matt 2:15 we realize that Matthew uses the verse differently. First, the very fact that he draws upon this verse from Hosea is peculiar considering what was said above about the lack of any messianic reference in the original text. Second, Matthew’s interest is in the geographical name Egypt which leads him to read “from Egypt” locally,[35] whereas it is temporal in Hos 11:1. Third, Matthew uses the verse within a theme of geography, whereas the original theme in Hos 11:1f is Israel’s apostasy despite YHWH’s calling.[36] The following chart may help in showing the differences:

Hosea Matthew
Historical, no messianic reference Interpreted messianically
Preposition is temporal Preposition is local
Talks about Israel Talks about Jesus the Messiah
Theme is apostasy and judgement Theme is geographical


It raises questions regarding Matthew’s hermeneutic. Is his interpretation faithful to Scripture or does it display randomness and distortion of it? Does the true meaning of prophecy lie in the prophetic text or in the experience of readers who apply it to new historical situations? Is it legitimate to interpret a text with no regard to the original context? Does fulfilment merely consist in allusion and hence rest on the imagination of the Gospel writer alone? And maybe most importantly, if canonical Matthew could treat the text this way, does it legitimize the same methodology in contemporary exegesis?

Conservative scholars have tried to get around the problem by insisting on a literal meaning (sensus literalis) of the text.[37] The fulfilment is thus strictly predictive, that is, the original meaning of the OT prophecy is a prediction about what eventually came to pass. The problem with such a view is that it forces a rationalistic Enlightenment hermeneutic on the text. It reduces truth to strictly historical truth, which in this passage is a violation of the author-intended meaning.[38] Others have argued that the prophetic text contained a hidden meaning (sensus plenior), which was divinely intended from the beginning but unknown to both the prophet and his audience.[39] It is likewise an admirable attempt to defend the integrity of Scripture, but it makes divine revelation through prophecy essentially unintelligible. One may ask if revelation which is not understood by humans – indeed not intended to be understood – is revelation at all. Furthermore, it opens up for arbitrary interpretations of Scripture because the meaning of a text rests on the perfect interpreter and hence completely detaches meaning from the text.

Much more convincing is the typological interpretation, which sees Jesus as a type of Israel.[40] The problem, of course, is that the fulfilment does not occur until v.21 when the return from Egypt to Israel takes place. However, the point of contact is the geographical name Egypt. The fact that Egypt is mentioned already in this pericope (v.13-15) makes it natural to quote the reference to Egypt already here. Furthermore, the pericope clearly says that the exodus is temporary, which means that the pericope itself contains the event of the return through the anticipation of it. The question is therefore not so much whether the correspondence between ancient type and contemporary fulfilment is fully accurate, but whether the quotation successfully communicates the intended meaning.

The purpose of the formula quotation is not to show the correspondence between prediction and fulfilment but to use Scripture to draw upon themes that Matthew found relevant to the life of Jesus. The fulfilment should therefore be understood more as an application, an understanding of the text in light of Matthew’s own time.

Exegesis of Matthew 2:23

The story of the settlement in Nazareth and the formula quotation concluding the pericope makes perfect sense in its Matthean context. Jesus ended up in Nazareth due to historical circumstances, but Matthew shows that God guided the events so that all took place according to his divine will as foretold by the prophets that he would be called a Nazarene. The only problem is that no prophetic text exists that matches the quotation in Mt 2:23. Indeed, the town of Nazareth is not even mentioned in the OT. This fact makes the verse one of the most controversial in the whole of Matthew’s Gospel. The question is which prophecy Matthew had in mind. Maybe he simply made it up or he was referring to a prophecy now lost.[41] Most scholars have been divided between three proposed solutions, which we will evaluate in turn.

Many have seen a reference to the Hebrew root of the word for ‘branch’ (נצר, nzr) in Is 11:1.[42] This view is favoured by the fact that Is 11 is in fact a messianic text and that it emphasizes the descent from David, which was so thoroughly spelled out in Matthew’s first chapter.[43] However, we have already seen from 2:15 that Matthew did not need a specifically messianic text before he dared to interpret it thus. Rather, he seems to place more weight on the historical events from the point of the fulfilment than on the prophetic text.

Moreover, this solution fails to explain several features in the text. Firstly, the formula before the quotation has irregularities when compared with the normal characteristics of the formulae. This one reads as follows: “so that what was said by the prophets might be fulfilled that …”[44] Firstly, we note that “prophets” is in the plural unlike the normal singular “prophet”, with (2:17) or without mention of the prophet’s name (1:22; 2:5, 15). Secondly, “saying”[45] as a marker of direct citation is absent. The replacement[46] can be read two ways, either introducing an indirect quotation or introducing a direct quotation (instead of a colon).[47] However, it still does not explain why the strong marker for the direct quotation is omitted, for Matthew could have retained both.[48] When compared to the rest of the formulae, 2:23 is markedly different in its generalizing and indirect character. Why would Matthew make this change if he referred to the single text in Is 11:1? Some have tried to solve the problem by adding secondary references,[49] but a quotation of Is 11:1 alone would have been sufficient to spell out the theological theme and was not in need of being enlightened by other texts.

Secondly, it is unlikely that Matthew would transliterate a Hebrew term into Greek without giving any explanation of the meaning (cf. ‘Jesus’ in 1:21; ‘Immanuel’ in 1:23). Even if Matthew’s setting was trilingual and many of his Jewish readers would have known at least some Hebrew, the wordplay would not have been obvious from the Greek text alone. We therefore conclude that the evidence for seeing a reference to Is 11:1 is insufficient.

Because of the indefinite nature of the formula others have denied reference to a single text altogether.[50] Thus instead of looking for a textual referent one should look for thematic connections.[51] The meaning of the fulfilment is therefore the disdain for Nazareth in contemporary society (John 1:46), which is projected backwards into Scripture where it finds support in the prophecies talking about the despised Messiah (Zech 11-13; Is 52-53). Jesus fulfilled prophecy when he moved to insignificant Nazareth because the place name contained the essence of what kind of Messiah he would become. Thus France can say that “[i]t is Nazareth’s very absence from the Old Testament which makes it a fitting fulfilment of this Old Testament theme.”[52]

This solution certainly has something to it. ‘Nazarenes’ was a well-known designation for Jesus’ followers before they came to be known as Christians (Acts 11:26; 24:5). Also Jesus himself was identified by his home town (Matt 21:11).[53] And as we have seen, the town was known to be provincial, as was Galilee generally, so that people certainly did not expect the kingly Messiah to come from there.

But what this solution has to offer when it comes to the fulfilment in the historical events, it lacks in the connection to the prophetic texts. Why would Matthew “quote” such an obscure group of texts when all the meaning is already contained in the reference to Nazareth? The Scriptural quotations normally add some theological meaning to the text (e.g. that Jesus is the new Israel), but in this instance Matthew would not have left out any theological theme had he omitted the quotation. We therefore seem to lack a referent in order to make the intended meaning of the pericope complete. There is simply no good reason for Matthew to include the formula quotation.

It is therefore most likely that Matthew had something different in mind, namely a reference to Judg 13:5.[54] Consider the different textual variants below:

MT Judg 13:5: כִּי־נְזִיר אֱלֹהִים יִהְיֶה הַנַּעַר מִן־הַבָּטֶן (“for the boy will be a nazir of God from the womb”).

LXXa[55] Judg 13:5: ὅτι ἡγιασμένον ναζιραῖον ἔσται τῷ θεῷ τὸ παιδάριον ἐκ τῆς γαστρός (for the boy will be consecrated [‘holied’] as a nazirite to God from the womb”).

LXXb Judg 13:5: ὅτι ναζιρ θεοῦ ἔσται τὸ παιδάριον ἀπὸ τῆς κοιλίας (“for the boy will be God’s nazir from the womb”).

LXXa Judg 16:17: ὅτι ναζιραῖος θεοῦ ἐγώ εἰμι ἐκ κοιλίας μητρός μου (“for I am God’s nazirite from my mother’s womb”).

LXXb Judg 16:17: ὅτι ἅγιος θεοῦ ἐγώ εἰμι ἀπὸ κοιλίας μητρός μου (“for I am God’s holy one from my mothers womb”).


We note that the Hebrew נזיר (nzr) has been transliterated directly into Greek nazir in LXXb Judg 13:5 whereas the same version translates with “the holy one” in Judg 16:17. LXXa has in both Judg 13:5 and 16:17 a transliteration with a Greek ending (naziraios). Comparing with Matthew’s nazoraios in 2:23 we realize that only one vowel distinguishes the two. The wordplay would have been less obvious in Hebrew but makes perfect sense in Greek without any additional explanation. The LXX rendering shows us that the transliterated term Nazirite known in Greek long before Jesus’ days. It is therefore likely that any Greek speaking Jew would have recognized the allusion.

The presence of the verb (κληθήσεται: “he will be called”) in Matt 2:23 has led some commentators to deny the Judg 13:5 passage as being the right background text. They argue that there is no explanation for Matthew’s alteration from “he will be” (ἔσται) to “he will be called”. However, the problem is not as weighty as it appears. Four possible solutions could be proposed. Firstly, the verb “to call” could have been derived from a parallel text such as Is 7:14[56] or Is 4:3.[57] The latter has more weight due to the textual link between the “holy one” in Judg 13:5 and the remnant who “will be called holy” in Is 4:3. The fact that the Nazirite is rendered in two different ways (“Naziraios” and “holy one”) in LXXa,b Judg 16:17 provides the textual evidence for seeing the two texts together. Secondly, we have noted that the LXX tradition was not yet standardized, which means that is it not totally unlikely that Matthew could have had a revised LXX in possession, which read κληθήσεται in Judg 13:5.[58] Thirdly, in line with Gundry’s argument about the text-form of Matthew’s quotations, we could simply see it as Matthew Targumizing over the verse.[59] Fourthly and most possible, we have noted that Matthew by means of the preceding formula gave the impression that this is not a full quotation but rather an allusion. It is therefore not necessary to find verbal links apart from the single word Nazoraios.

Other arguments against the reference to Judg 13:5 have been put forward. Some have contended that Jesus was not a Nazirite but “a glutton and a drunkard, a friend of tax collectors and sinners” (Mt 11:19).[60] However, this misses the point because the rationale for including the reference was not the original context of the prophecy but the contemporary geography of Jesus. The point is therefore not that Jesus was a Nazirite, but rather that Matthew found evidence in Scripture that the Messiah came from Nazareth.

Others have argued that there is no explanation for the vowel change between Naziraios and Nazoraios. However, any textual objection carries little weight here because the link is due to hermeneutics rather than to textual association. Matthew had Judg 13:5 in mind because his hermeneutic allowed him to make free allusive connections based on wordplay, not because he found the exact same text in one of the text-forms available to him.

We therefore conclude on the formula quotation in Matt 2:23 that Matthew intended an allusion to Judg 13:5. He might have seen a typological link between the boy consecrated to God in the book of Judges and Jesus as the holy one and God’s own son. However, the immediate rationale was the historical geography of the Messiah. Matthew’s primary aim was to record the events as he knew them and to present this historical fact as fulfillment of prophecy, not to draw on themes from Scripture.

Geography in Matthew 2

We have seen that the intended link between Matthew’s story and prophecy is the geographical references. As Matthew wrote his Gospel, the historical birth in Bethlehem reminded him of a prophecy mentioning Bethlehem (Mic 5:1); the flight to Egypt reminded him that the son of God already once came out of Egypt (Hos 11:1); the disaster in Bethlehem reminded him of the weeping of Rachel (who was associated with Ephratah/Bethlehem; Jer 31:15); and finally, his settlement in Nazareth was alluded to by a wordplay on the Nazirite (Judg 13:5) because no mention of the place name Nazareth was to be found in Scripture.

Considering the role of geography in traditional societies, it should not surprise us that following a theme of descent (Matt 1) we find a geographical theme (Matt 2). In traditional societies a person’s identity is closely linked to the geographical origin because the geographical name reveals family tree. Thus it required an explanation that the son of David ended up being known as a Nazarean and that his followers became known as the Nazarene sect. There is no need to see this apologetically, for this was simply the first question to ask in traditional societies when a new person was introduced. People in modern Western societies have their identity in what they do (their job), thus one of the first questions people want to know when meeting a new person is what this person does. However, in traditional societies, one of the first questions people ask when meeting a new person is where he comes from. The identity lies in the geographical origin. Thus there was no need to see Matt 2 as an apologetic for Jesus’ affiliation with Nazareth, for that was simply the first information people would ask for. It therefore follows entirely naturally from the matters of descent as treated in Matt 1.

However, we cannot deny that the place names carried allusive meaning. Bethlehem was the royal city where the Messiah was to be born; Egypt was the prime reference to the Exodus; and Nazareth at the edge of Israel’s territory contained by its very absence from Scripture a significant negative connotation. It is likely that we in the place names shall read a theme of inclusiveness against the predominant Judean exclusivity. If an Israel-Jesus typology is intended in 2:15 – and I believe it is – then Jesus is actually a representative of the whole of Israel, from the greatest lineage of David to the humble Nazarene origin and from the proudly sophisticated urban center to the peasant rural area. This entire nation bears the promises of God through his sovereign election when he called them out of Egypt. Jesus, being the center of Scriptural prophecy, gathered under him all of Israel as was told by the prophets. By so doing he opposed a prevailing tribalist Jewish mindset.

How, then, should we understand the relation between prophecy and fulfilment in Matthew? In order to answer this question we will need to think in terms of contemporary Jewish hermeneutic where prophecy entailed much more than what can be explained by a ‘prediction-come to pass’ way of understanding. The pesher tradition understood the prophetic utterance as secondary to the fulfilment by historical events, as has been emphasised by Stendahl. The Targums were similarly actualizing interpretations of ancient texts by means of the process of translation. Again, typology was a well-known method of understanding the texts in light of events taking place at the time of the interpreter. These methods all think about Scripture as revelation with a referent in historical time, but the universal meaning is outside of time. Matthew understood Scripture in light of the historical events as he knew them and interpreted Jesus as the fulfilment of the whole of Scripture.

Modern interpreters may question some of the hermeneutical principles, especially of later developments in rabbinic interpretation.[61] However, Matthew’s gospel is a contextual document as is the rest of Scripture, and there is every reason to believe that Matthew communicated competently to his contemporaries even if modern readers fail to recognize his methods. The core of Matthew’s infancy narrative is historical event, which was subsequently put into a theological framework and given theological significance by Matthew. The result is theologically interpreted history.


We have tried to establish a framework for understanding Matthew’s use of Scripture in relation to the fulfilment in Jesus. This framework has to be contextual in order to adequately explain the author-intended meaning. Jewish hermeneutical categories (particularly pesher, Targumic translation, and typology) – though we should be careful not to limit ourselves to tight categorizations – are helpful in understanding the hermeneutical environment at the time of Jesus and Matthew.

The formula quotation in Matt 2:15 is a quotation of Hos 11:1 in a text-form unknown to us, which contained a rendering useful for the purposes of Matthew (“my son” in the singular). The primary focus for Matthew is the geographical reference which Matthew picked up because he saw a parallel in Jesus’ early life. The quotation creates an Israel-Jesus typology, which communicates that Jesus is the fulfilment of the calling of Israel in its totality.

The formula quotation in Matt 2:23 is likewise primarily a geographical allusion (not a quotation per se), which refers to the Nazirite in LXXa Judg 13:5. The original context of the text is at the most secondary and possibly not in Matthew’s mind at all. The point is to show that the whole of Israel, from Bethlehem to Nazareth, is gathered under Jesus the Messiah as was prophesied in Scripture.

This interpretation can seem unconvincing to a modern reader, but Matthew did not write for a modern rationalistic Enlightenment audience. He wrote for a traditional Jewish messianic sect in the first century. Drawing upon conventional hermeneutical methods he established Jesus’ geographical origin in line with traditional ways of presenting a newly introduced person’s identity. He did so in no extraordinary way and for no extraordinary reasons other than to show that this is the Messiah who is the fulfilment of Scripture.



Brown, R. E. The Birth of the Messiah: A Commentary on the Infancy Narratives in the Gospels of Matthew and Luke. NY: Doubleday, 1993 (2nd ed.).

Chrysostom. Homilies on the Gospel of St. Matthew. In P. Schaff (ed.): Nicene and Post-Nicene Fathers, Vol. 10. Edinburgh: T & T Clark, 1888.

Davies, W. D. & D. C. Allison. A Critical and Exegetical Commentary on The Gospel According to Saint Matthew: Introduction and Commentary on Matthew I-VII. Vol I. International Critical Commentary. Edinburgh: T. & T. Clark, 1988.

Enuwosa, J. The Soteriological Significance of Matthew 2:15 in his Use of Hosea 11:1 from an African Perspective. Africa Theological Journal, Vol. 24, no. 1 (2001), pp.39-52. 

France, R. T. The Formula-Quotations of Matthew 2 and the Problem of Communication. In New Testament Studies, Vol. 27 (1981), pp.233-251.

Geiger, G. Falsche Zitate bei Matthäusund Lukas. In C. M. Tuckett (ed.): The Scriptures in the Gospels. Leuven: Leuven University Press, 1997.

Gundry, R. H. The Use of the Old Testament in St. Matthew’s Gospel with Special Reference to the Messianic Hope. Leiden: Brill, 1975.

Harris, R. Testimonies. Cambridge University Press, 1916-20.

Howard, T. L. The use of Hosea 11:1 in Matthew 2:15: An Alternative Solution. In Bibliotheca Sacra, Vol. 143, no. 572 (October 1986), pp.314-328. 

LaSor, W. S. Prophecy, Inspiration, and Sensus Plenior. In Tyndale Bulletin, Vol. 29 (1978), pp.49-60.

Longenecker, R. Biblical Exegesis in the Apostolic Period. Grand Rapids, MI: Eerdmanns, 1999.

McCartney, D. G. & P. Enns. Matthew and Hosea: a response to John Sailhamer. Westminster Theological Journal, Vol. 63, no. 1 (Spr 2001), pp. 97-105.

McCasland, S. V. Matthew Twists the Scripture. In G. K. Beale (ed.): Essays on the Use of the Old Testament in the New: The Right Doctrine from the Wrong Texts. Grand Rapids, MI: Baker Books, 1994, pp.146-152.

Menken, M. J. J. The Sources of the Old Testament Quotation in Matthew 2:23. In Journal of Biblical Literature, Vol. 120, no. 3 (2001), pp. 451-468.

Menken, M. J. J. Matthew’s Bible: The Old Testament Text of the Evangelist. Leuven: Leuven University Press, 2004.

Pesch, R. ‘He will be Called a Nazorean’: Messianic Exegesis in Matthew 1-2. In C. A. Evans & W. R. Stegner (eds.): The Gospels and the Scriptures of Israel. JSNT Supplement Series 104. Sheffield: Sheffield Academic Press, 1994, pp.129-178.

Powery, E. B. Jesus Reads Scripture: The Function of Jesus’ Use of Scripture in the Synoptic Gospels. Biblical Interpretation Series Vol. 63. Leiden: Brill, 2003.

Prabhu, G. M. S. The Formula Quotations in the Infancy Narrative of Matthew: An Enquiry into the Tradition History of Mt 1-2. Analecta Biblica 63. Rome: Biblical Institute Press, 1976.

Sailhamer, J. Hosea 11:1 and Matthew 2:15. Westminster Theological Journal, Vol. 63, no. 1 (Spr 2001), pp.87-96.

Sanders, J. A. Nazoraios in Matthew 2.23. In C. A. Evans & W. R. Stegner (eds.): The Gospels and the Scriptures of Israel. JSNT Supplement Series 104. Sheffield: Sheffield Academic Press, 1994. pp. 116-128.

Stendahl, K. The School of St. Matthew and its Use of the Old Testament. Philadelphia: Fortress Press, 1968 (2nd ed.).


[1] ‘Formula quotation’ refers to a quotation from Scripture, which is introduced by a standard formula. This formula, with slight variations, has the following format: “… that it might be fulfilled what was spoken by the prophet, saying: …” (The 11 formula quotations in Matthew occur in Mt 1:23; 2:15, 18, 23; 3:3; 4:15; 8:17; 12:18; 13:35; 21:5; 27:9).

[2] This paper has been limited to these two examples because of the vastness of the subject. However, the result of our analysis will apply to the second half of the infancy narrative in particular and Matthew’s use of Scripture in general.

[3] LaSor (1978); Enuwosa (2001).

[4] Sailhamer (2001). Concerning the reference in Matt 2:15 to Egypt, he argues that Hosea (Hos 11:1) as well as Matthew (Matt 2:15) gave the sensus literalis (literal meaning) of the Exodus.

[5] Geiger (1997).

[6] McCasland (1994).

[7] Pesch (1994).

[8] Harris (1916-20).

[9] Gundry (1967), p.166.

[10] Stendahl (1968). The 1968 edition was a revision of his first work in 1954.

[11] The question of Sitz im Leben (life setting) of the biblical texts has been one of the central questions of historical-critical scholarship during the 20th century. The School of St. Matthew is Stendahl’s answer to that question, guided by his form-critical assumptions. Stendahl’s analysis offers great new insight, but we need not accept his theory of the School of St. Matthew in order to make use of his treatment of the pesher hermeneutic.

[12] Stendahl (1968), p.35. Stendahl lists Brownlee’s 13 principles describing the pesher exegetical method. Of these can be mentioned: “Everything the prophet wrote has a veiled, eschatological meaning”; “A textual variant, i.e. a different reading from the one cited, may also assist in interpretation”; “sometimes the prophet veiled his message by writing one word instead of another, the interpreter being able to recover the prophet’s meaning by … the substitution of similar letters for one or more of the letters in the word of the biblical text” (see p.191-192). Stendahl’s categorization of the pesher method as a subcategory alongside the halakic (concerned with conduct) and haggadic (concerned with edification) methods under the category of midrash has been criticised by Longenecker. If pesher be considered midrashic, then a qualification needs to be made. Whereas rabbinic midrash exegesis said “That has relevance to This”, the pesher exegesis said “This is That”. Thus if applying the term midrash for the pesher method, he advises that we distinguish between the Qumranic charismatic midrash (pesher) and rabbinical scholastic midrash (1975, p.43-44).

[13] Prabhu (1976), p.74; Howard (1986), p.318f; Gundry (1967), p.155f.

[14] Brown (1993), p.102, n.13.

[15] Gundry (1967), p.2.

[16] The Markan quotations (which include the synoptic parallels) are traditionally considered septuagintal in text-form whereas Matthew’s formula quotations are mixed text-forms. Thus by including allusive quotations Gundry argued that the Septuagintal text-form was the exception rather than the rule.

[17] The Targums are translations of Scripture into Aramaic. The translations paraphrase rather than translate literally, which gives them a significant interpretive element. Gundry comments on the relation between pesher and Targum: “What is midrash-pesher but the targumic method applied to prophecy believed to be fulfilled or on the verge of fulfilment in an eschatological situation? It is the fulfilment-motif which gives NT interpretation of the OT affinity to Qumran pesher. But in the mechanics of the text-handling, both rest on the targumic method” (1967, p.174).

[18] Gundry does mention typology as a commonly used hermeneutic principle along with prophecies having “direct applicability” to Jesus (1967, p.208f). While Mt 2:15 may be an example of typology, neither of these two hermeneutical principles fully explain the formula quotation in Mt 2:23.

[19] Prabhu (1976).

[20] Brown (1976).

[21] Brown (1976), p.32ff; Prabhu (1976), p.299.

[22] Pesch (2004), p.178.

[23] That is, transcending history, as opposed to un-historical (incorrect history).

[24] France (1981), p.236.

[25] MT is abbreviation for “Masoretic text”.

[26] LXX is the abbreviation for the “Septuagint”.

[27] France notes that the markedness of a quotation sometimes lies precisely in the fact that it has been altered: “A ‘twisted’ quotation can communicate very special nuances of meaning, just as twisted” (1981, p.243).

[28] Gundry (1967), p.174ff.

[29]Following the genealogy, the infancy narrative contains five pericopes, each containing a formula quotation: The birth and naming of Jesus (1:18-25); the Magi (2:1-12); the flight to Egypt (2:13-15); the slaughter in Bethlehem (2:16-18); the settlement in Nazareth (2:19-23).

[30] Gundry (1967), p.93; Prabhu (1976), p.216; Geiger (1997), p.479-480.

[31] Menken (2004).

[32] That is, in the material that is his own and not taken over from Mark (see Mt 19:1; 27:55). Even in some of the Markan material, Matthew changes the preposition ἐκ to ἀπὸ: Mk 1:10/Mt 3:16; Mk 6:14/Mt 14:2; Mk 9:25/Mt 17:18; Mk 13:1/Mt 24:1; Mk 13:25/Mt 24:29 (Menken 2004, p.139-140).

[33] Gundry (1967), p.93.

[34] The argument becomes clear if we consider the absurdity of a local understanding of the reiterated calling in v.2: “the more I called them out of Egypt, the more they went from me”. It is evident that the turning away from YHWH concerned the apostasy during the history of Israel. Prabhu (1976, p.216) rightly points to Hos 11:11; 12:14 as clear examples of Hosea’s local use of the preposition, but it does not necessitate the same use in v.1. Hos 12:10 is an unambiguous example in Hosea of the temporal use (Gundry, 1967, p.93, n.5).

[35] Gundry believes that the preposition is both local and temporal both in Matthew and Hosea. However, this does not adequately take the contexts into account. The context in Hosea requires a temporal rendering while the context in Matthew is geographical.

[36] Howard suggests a comparison of the exodus from Herod with the exodus from Pharaoh as a solution to the fact that strictly speaking Matt 2:15 was only fulfilled in v.21 where Jesus moved out of Egypt (1986, p.321; cf. McCartney & Enns 2001, p.103). Thus he believes the persecution in Egypt under Pharaoh parallels the persecution of Jesus in Israel under Herod. However, paralleling Egypt and Israel (rather than Egypt in Hos 11:1 and Egypt in Mt 2:13-14) ignores the fact that Matthew’s interest throughout chapter two is geographical rather thematic. Matthew quoted the verse because Jesus ended up in Egypt (which reminded him of the reference to Egypt in Hos 11:1), not because the flight reminded him of the exodus-motif.

[37] Sailhamer (2001).

[38] McCartney & Enns (2001), p.100.

[39] Enuwosa (2001).

[40] France (1981), p.243; Powery (2003), p.98. Howard prefers the term “analogical correspondence” because typology tends to assume a deeper level of meaning in the original text similar to sensus plenior (1986, p.320).

[41] Chrysostom, Homilies on Matthew 9:6.

[42] So Stendahl (1968); p.199; Gundry (1967), p.196; Pesch (1994), p.174;

[43] Mt 1 has shown considerable interest in proving that Jesus was the son of David, first by establishing the Davidic genealogy (1:1-17) and then by emphasizing the role of Joseph in the naming process (1:18-25).

[44] The Greek has: ὅπως πληρωθῇ τὸ ῥηθὲν διὰ τῶν προφητῶν ὅτι.

[45] Mt 1:22; 2:15, 17 have ”by the prophet, saying:” (διὰ τοῦ προφήτου λέγοντος).

[46] Here in 2:23 the quotation is introduced by ὅτι.

[47] The latter is the recitative use of ὅτι. Menken’s suggestion that the ὅτι is part of the quotation is another possibility but helps little in explaining the origin of the quotation (2001, p.461).

[48] Thus using the same construction as e.g. Mt 9:18; 10:7, etc.

[49] Pesch (1994), p.175.

[50] France (1981), p.247. Powery (2003), p.100; Prabhu (1976): ”[T]he passage introduced does not really quote a definite OT text but at most alludes to one” (p.58). This allusion he believes is Judg 13:5.

[51] France (1981), p.247.

[52] France (1981), p.248.

[53] The Book of Acts uses Nazareth to identify Jesus 7 times, 2 times also with “Christ” (3:6; 4:10).

[54] Prabhu (1976), p.205f; Sanders (1994), p.122; Menken (2001), p.461.

[55] The LXX has in some places two parallel text families. They are labelled by the letter a and b.

[56] So Menken (2001), p.464.

[57] Davies & Allison (1988), p.277. This also offers an explanation for the plural “prophets” in the formula. However, the plural is easily explained by the reference to Judg 13:5 only, for the Book of Judges was not known as the product of a single prophetic author but rather as the former prophetic books. On dealing with this question, Menken concludes: “If Matthew wished to quote from Judg 13:5, 7, that is, from the former prophets, it was only natural for him to write ‘by the prophets,’ in the plural” (2001, p.467).

[58] This solution, however, must be considered speculative. However, the fluidity of the LXX tradition has become an increasingly recognized view. Presently, we simply do not have evidence enough to either accept or reject the possibility.

[59] Gundry (1967), p.172ff.

[60] France (1981), p.248.

[61] Longenecker (1975), p.35-36.


Entry filed under: Posts in English. Tags: , , , , , , , , , , , , , .

John Mbiti kommer til NEGST I kirke i World Harvest International v/ Apostel Ssuna

7 kommentarer Add your own

  • 1. German Note  |  juni 4, 2009 kl. 12:52 pm

    Hi Soren!

    Had some spare time and could not find the texts I should have read for my “Covenant Seminar” so I thought about looking at your page again. Have not completted reading it, but so far it’s interesting.
    Where you jump with two questions, immediately to the historical Faktizität:-) of Mt 2, is for me a little fast. but … whatever

    I need to run to “Mensa” to get a bit to eat and then I will read a little bit from the “größten Christozentriker, den es je gab” (Zinzendorf). I think it’s funny that the quote comes from Barth, who talks always like this:”nicht was, sondern wer, nicht ein Objekt, sondern ein Subjekt, es geht immer nur um eins, nein nicht um eins, sondern um einen, um ihn: Christus” (ala Barth free Wiedergabe from me).

    More later


  • 2. Søren Dalsgaard  |  juni 8, 2009 kl. 8:29 am

    Ahhh … Zinzendorf and Barth. You are back to European theology can I see. Sounds great!
    What is your covenant seminar about? Covenant theology?

  • 3. Christian  |  september 3, 2010 kl. 1:01 pm


    How’s life?
    Covenant Seminar was about all the different covenant usings in the OT and in oriental literature.
    AND I had already visited one in Germany, but I wanted to do the same again:-) to understand it:-)
    still am a little clueless with some things, but
    I am finished my paper on “the new covenant in Jer 31” after about four years now:-) it’s ok. Would give it to you for reading, but it’s in German.

    But here why I write.
    I’m thinking about making holidays in Kiel. TO Denmark it’s not very far from there. So should we hook up. Are you two around there somewhere? Have you enjoyed the WORLDCUP:-) in SA?
    I am pretty unplanned again and I will be in Kiel probably on Monday. THIS Monday:-) So let’s see. I might be there for a week?
    and then my final year will come. learning one year for my final exame. YEAAAAAAAAAAAAAAAAAAAh

    greetings to Charlotte

  • 4. Tom Sims  |  maj 3, 2012 kl. 4:25 pm

    Dear Soren,
    Did you ever return to this research, or finish the article on the meaning of the word ‘fulfill’? If you did I would love to see your results, I am doing a translation of Mathew and am very interested in your work.

    in Messiah,

    Tom Sims

  • 5. Søren Dalsgaard  |  maj 3, 2012 kl. 8:18 pm

    Dear Tom. I have sent you an email. Blessings, Søren.

  • 6. BlurBlue-Note  |  februar 18, 2013 kl. 11:18 pm

    great points altogether, you simply received a logo new target audience.
    What may you recommend about your post that you
    simply made a few days before? Any sure?

  • 7. Rachel  |  april 10, 2013 kl. 5:45 am

    Hi, just wanted to say, I liked this article. It was practical.
    Keep on posting!


Skriv et svar

Udfyld dine oplysninger nedenfor eller klik på et ikon for at logge ind: Logo

Du kommenterer med din konto. Log Out / Skift )

Twitter picture

Du kommenterer med din Twitter konto. Log Out / Skift )

Facebook photo

Du kommenterer med din Facebook konto. Log Out / Skift )

Google+ photo

Du kommenterer med din Google+ konto. Log Out / Skift )

Connecting to %s

Trackback this post  |  Subscribe to the comments via RSS Feed

Nye indlæg

%d bloggers like this: