One of the first things that got me interested in studying language was a doctoral dissertation by Dr. Eric Svendsen on the topic of Mary in the New Testament and Roman Catholicism. In this dissertation, Svendsen discusses the discusses the traditional Roman Catholic interpretation of Matthew 1:25 and the brothers of Jesus which goes back to the discussion on the perpetual virginity between Jerome and Helvidius. The debate on Matthew 1:25 centered on the word “until:”
Matthew 1:25 but kept her a virgin until she gave birth to a Son; and he called His name Jesus.
Helvidius stressed that the word “until” here indicates the notion that Joseph and Mary engaged in normal sexual relations after Jesus was born. Jerome [whose arguments are still followed by Roman Catholic apologists] argued that the word “until” does not necessarily indicate that Mary had normal sexual relations with Joseph. There are several passages that could be used in this regard, but, just for example, take the following in the LXX:
2 Samuel 6:23 And Melchol the daughter of Saul had no child till the day of her death.
Clearly, in this instance, no one is going to suggest that Melchol started having children after she was dead.
Svendsen’s contribution was to distinguish between different phrases that are translated “until” in Greek. The Greek term heos alone can be used, as can the phrases heos an, heos hou, etc. What Svendsen pointed out is that the Greek here is heos hou, and in the New Testament as well as in the literature 200 years surrounding the birth of Christ, the phrase heos hou means “until a specified time [but not after].”
Roman Catholic and Eastern Orthodox apologists reacted to this thesis in various ways. Some people absolutely did not know what they were talking about, and ignored the fact that Svendsen’s thesis is based upon Saussure’s distinction between diachronic and synchronic study of language. Hence, they couldn’t understand why Svendsen rejected the uses in the LXX as being relevant. In fact, many people misrepresented his views as saying that Svendsen did not view the examination of the usage of the term in the LXX as relevant to the meaning in the NT, and even misrepresented his view to scholars garnering alleged rebukes of him by famous protestant scholars. Of course, the reason why the usages in the LXX did not affect his thesis is because he demonstrated a disconnect between the way the phrase was used in the LXX and the way it was used in the time period of the NT thus indicating diachronic semantic variance. It is this diachronic variance that allowed him to conclude that the LXX uses did not affect his argument.
Others took a more reasonable tactic, and pointed out that there is other literature that is of questionable dating that may date from the 200 years surrounding the birth of Christ which contain uses of heos hou where the action of the main clause is not terminated. Svendsen pointed out that even if such dating is correct, one still has to take into account archaizing tendencies on the part of some authors, and, indeed, if what he is saying is correct, then the usage of heos hou in this way may indicate that the date of these works are not to be placed in the 200 years surrounding the time of Christ.
I still largely agree with what Svendsen wrote. Even if one could come up with instances of heos hou in the two centuries surrounding the birth of Christ, it would still make Jerome’s interpretation highly unlikely. As Svendsen points out, even in the LXX, the meaning of heos hou as “until [and continuing]” is quite rare. Thus, although Svendsen’s thesis does not settle the issue absolutely, it most certainly does make the Roman Catholic interpretation very highly unlikely.
However, as I have been studying pragmatics, I have wondered if Svendsen’s argument is incomplete. The reality is, Svendsen, in order for his thesis to even be written, had to assume that there was a way in which we can tell whether the action of the main clause is terminated by the action of the subordinate clause. It is this very methodology that he used to categorize all of these uses of heos, heos hou, etc. Likewise, Roman Catholics, when they were looking for counterexamples, had to have a way of telling whether something is a counterexample or not. This suggests that there is a way in which this issue can be settled by simply relying upon normal rules of communication. It is here that I believe that Gricean and Neo-Gricean implicature can help us.
A little background is in order here. Grice’s theories of implicature developed out of a desire to understand what goes on in conversation. He started with certain things that speakers and listeners assume when they are interacting. The main principle of Gricean and Neo-Gricean theories is what is called the “Co-operative principle.” This is summarized by Huang as
Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged[1].
Grice then lists four maxims which help us accomplish this:
1. Quality-Make sure that your contribution is something you believe to be true, and not something you believe is false or which you are unsure of.
2. Quantity-Make sure you are as informative as is required, and not more informative than is required.
3. Relation-Make sure your contribution is relevant.
4. Manner-Be clear by avoiding ambiguity, prolixity, and disorder[2].
In Neo-Gricean theories, some of these principles were collapsed into one. For example, in Horn’s version, all of these maxims are divided into Quantity [Q] and relation [R] maxims, while in Levinson’s Neo-Gricean theory, they were divided into Quantity [Q], Informativeness [I], and Manner [M].
These principles help us to avoid a major problem in language. Consider the following example given by Huang[3]:
John has had nine girlfriends.
a. John has had at least nine girlfriends.
b. John has had exactly nine girlfriends.
The sentence “John has had nine girlfriends” is subject to the interpretation found in both a and b. We could treat this sentence as semantically ambiguous, but the problem is that such an interpretation runs the risk of falling prey to Occam’s razor[4]. The difficulty also is that there are other words that are likewise subject to this kind of interpretation such as “all” or “some.” If someone says, “Some of the candy in the bag is purple,” in terms of strict semantics, this does not rule out the possibility that *all* the candy in the bag is purple. If all the candy in the bag is purple, then some of the candy in the bag is purple. This is a major problem since, as Huang suggests, it would turn a dictionary into an exercise in proliferation[5]. The same analysis can be give to our text:
Joseph kept her a virgin until she gave birth to a son.
a. Joseph kept her a virgin at least until she gave birth to a son.
b. Joseph kept her a virgin only until she gave birth to a son.
Huang, in discussing this problem, suggests that this problem can be solved by a division of labor between semantics and pragmatics. He suggests that, semantically, we can assign the meaning “at least” while getting to “only” or “exactly” through the use of pragmatics, and, imparticular, conversational implicature[6]. The reason has to do with Grice’s Quantity maxim, namely, that a speaker must not be more informative than is required. Thus, one can say that, both in the case of the number 9, and the word “until,” the reason the speaker doesn’t say “exactly nine” or “only until” is that, to do so, would be superfluous. We assume that speakers mean “only” or “exactly” precisely because we assume that they are giving us sufficient information. The same thing goes for the word “some.” If someone says, “Some of the candy in the bag is purple,” we assume that they are giving us sufficient information, and we likewise conclude that only some and not all the candy in the bag is purple.
I had this experience at work recently. A customer and I were looking to see if a particular camera came with a SD card, and we found the word “contents” and then a list [camera, charger, instruction manual, etc.] but we did not find “SD card” on the list. Hence, she went over, and started looking at the SD cards. Why? Because she assumed that the list contained *only* the things written under the “contents.” She assumed this based upon the fact that the person who wrote the list was being sufficient in his communication.
Hence, we would have, by Q-Implicature:
He kept her a virgin until she gave birth to a son.
+>He kept her a virgin only until she gave birth to a son.
So, we have accounted for how protestants read this passage. Now, why is it that Roman Catholics can point to examples of “until” where it clearly does not mean “only until?” The answer is very simply that conversational implicatures can be defeated. For example, Huang calls our attention to this sentence[7]:
His wife is often complaining.
+>His wife is not always complaining.
However, consider what would happen add another sentence:
His wife is often complaining. In fact, she complains all of the time.
~+>His wife is not always complaining.
Normally, we assume that when a person uses the term “often” that they are giving us sufficient information, and thus, they do not mean to suggest that his wife is always complaining. In this case, the semantic entailment of the second sentences [she complains all the time] defeats the implicature. Implicature can also be defeated by background assumptions and context. Consider the example we looked at earlier:
And Melchol the daughter of Saul had no child till the day of her death.
~+>Melchol, the daughter of Saul, had no child only until the day of her death.
The reason why this implicature is defeated is because of our background knowledge of the way in which childbearing works. Childbearing only occurs when a person is alive, not when they are dead.
The point is that we assume that the author is giving sufficient information [only until] unless we have semantic entailment, background assumptions, or something in the context that contradict the implicature. This is why I believe Svendsen was able to categorize all of the examples of the various phrases.
Also, this will work regardless of the Greek phrase used. Consider the Greek phrase heos an which is notorious for being used when there is no termination in the action of the main verb:
Matthew 2:13 Now when they had departed, behold, an angel of the Lord appeared to Joseph in a dream and said, “Rise, take the child and his mother, and flee to Egypt, and remain there until I tell you, for Herod is about to search for the child, to destroy him.”
Remain there until I tell you.
+>Remain there only until I tell you.
This interpretation is born out by the fact that, after Herod dies, the angel tells them to return.
Mark 6:10 And he said to them, “Whenever you enter a house, stay there until you depart from there.
Stay there until you depart from there.
+>Stay there only until you depart from there.
This interpretation is born out by the fact that it would be foolish to assert that, once they depart from that place, they are going to be staying in the same place. However, consider instances in which heos an very clearly cannot mean “only until.”
Matthew 22:44 “‘The Lord said to my Lord, Sit at my right hand, until I put your enemies under your feet’?
Sit at my right hand, until I put your enemies under your feet
~+>sit at my right hand only until I put your enemies under your feet.
Why does this implicature not follow? It does not follow because of the context of the Psalm [110:4] which says that he is a priest forever, but also because of our background knowledge that both the New Testament present the kingdom of Christ as something which endures forever. Hence, it cannot be that the honor which Christ receives lasts only until his enemies are put under his feet.
Let us now return to the implicature I proposed earlier:
He kept her a virgin until she gave birth to a son.
+>He kept her a virgin only until she gave birth to a son.
In order for the Q-implicature to be defeated, the Roman Catholic or Eastern Orthodox apologist must show that there is something that the passage entails, that we assume as background, or that is in the context that would defeat this implicature. If they cannot do that, then, by the normal assumption that speakers are giving us sufficient information, we can assume that Matthew is contradicting the Roman Catholic doctrine of the perpetual virginity.
However, what is worse is that I believe the context actually confirms the Q-implicature I have noted earlier. Consider the beginning of this pericope:
Matthew 1:18 Now the birth of Jesus Christ was as follows. When His mother Mary had been betrothed to Joseph, before they came together she was found to be with child by the Holy Spirit.
The crucial phrase is “before they came together she was found to be with child by the Holy Spirit.” Words like “before” trigger what is called a “presupposition.” To see this, simply use the word “before” in another sentence:
Before I went to the store, I mowed the lawn.
>>I went to the store.
Before we saw the eclipse, we saw a spectacular meteor shower.
>>We saw the eclipse.
Before we saw the aurora borialis, we looked at the newspaper to see if the conditions were right for it to occur.
>>We saw the aurora borialis.
Before I caught the ball, I slipped and fell.
>>I caught the ball.
I could continue, but I think you get the point. Hence, going back to Matthew 1:18, we would have:
before they came together she was found to be with child by the Holy Spirit.
>>They came together.
Roman Catholic and Eastern Orthodox apologists have not so much rejected this conclusion as they have the meaning of the term “come together” as having any relevance to sexual relations. They suggest that it is not meant to be taken in a sexual fashion. The problem is that, if you take the phrase “they came together” in a sexual fashion, it provides a nice inclusio between verses 18 and 25 as both would be clear statements of the virgin birth. Also, if we take this to simply be referring coming together in a living arrangement, then it would have virtually no relation to Mary being found to be with child. That would be a problem both before they came into the living arrangement, as well as after they came into the living arrangement. Not only that, but the point seems to be that Joseph knows that he is not the father of the child, and that is why he wants to put her away. That would fit nicely with “before they had sexual relations, she was found to be with child by the Holy Spirit.”
Hence, the real problem with the argument of those who hold to the perpetual virginity is, instead of arguing contextually, they have run off to other instances of “until,” which are not pragmatically parallel to Matthew 1:25. However, if the Q-implicature in Matthew 1:25 is to be defeated as it is in the texts they like to point to, they must show that some parallel exists, when it very clearly does not.
[1]Huang, Yan. Pragmatics. Oxford Textbooks in Linguistics. Oxford University Press. New York, NY. 2007. p.25
[2] ibid.
[3] ibid, p.7
[4] ibid.
[5] ibid.
[6] ibid, pgs. 7-8
[7] ibid, p.32
+> Conversationally Implicates
~+> Does Not Conversationally Implicate
>> Presupposes
