Skip to content

Much Ado about Shwa

April 11, 2012

Shwa (schwa) means “nothing.” In the Hebrew vowel system, shwa is the name for the most reduced, most indistinct vestage of a vowel. (We have borrowed the term for our symbol Ə). It is written underneath a consonant and consists of two dots, looking like a tiny colon (:). It happens to be the first vowel in the Bible, underneath the letter beth inbereshith, the first word of Genesis.

Understand that Moses never wrote a shwa. It is part of a system dating from about the eighth century A.D. in which the traditional pronunciation was preserved by diacritical marks, dots and dashes, which indicated phonetic details such as vowels, accents, pauses, punctuation etc. They are not letters as such–only the consonants are the original text–but they do carry great weight, as the scribes who employed them were remarkably meticulous.

The reason that I mention the shwa is that I need it in order to discuss an exegetical issue involving the first verse of the Bible. There exists what I will call the “alternative parsing” (AP), as opposed to what I will call the “standard parsing” (SP). Which option one chooses makes a difference in meaning and translation, as follows:

SP: In the beginning, God created the heavens and the earth (ESV, likewise KJV, NIV, etc., etc.)

AP: When God began to create heaven and earth (JPS Tanakh, likewise NEB, and YLT)

I’ll explain a little later how the AP works. It is preferred within a certain circle, though I do not think it is correct–or even has very much to commend it as an option. In this post I intend to explain why. This explanation is necessarily technical and I will address myself to those unfamiliar with Hebrew grammar, with apologies to those who do.

I began by explaining the shwa, because this “nothing” is the only thing on which the AP is based, that is the sole graphic mark which the proponents of the AP can point to in the text to indicate how they understand the syntax of the verse. From its presence under the prefixed proposition b- (“in”), they parse the noun reshith (“beginning”)as being in something known as the “construct state,” which I will now endeavor to explain.

Construct State

Imagine two nouns in a relation we might characterize as “X of Y.” Indeed English indicates this idea with by putting the preposition of between X and Y. Many other languages do likewise. In some cases we can  reverse the order and add a suffix to the Y element: Y’s X. Several languages have this pattern.

In Greek, this same relation is indicated by a marking on the Y element (putting the noun in the genitive case), though the order of the words is not fixed, e.g. he agape tou theou (“the love of God”).

In Hebrew, the same relation is indicated by forming a chain in the order X:Y with the X element marked by being lightened, reduced stress, reduced vowels (as far as possible), and some variations in the form of suffixes. This lightened form is the “construct state.” It is not limited to two elements, but can be X:Y:Z or more, with both X and Y in the construct state and only Z being in the default “absolute state.” Also the entire phrase is either definite or indefinite. If there is an article, it only occurs on the final element, and never present on a noun in theconstruct state.

A familiar example is the name of the town Bethlehem, which means “house of bread.” The noun meaning house in absolute state is bayit. In construct it becomes betLehem means “bread.” So bet-lehem, is “house of bread.”


It is absolutely factual that the prepositional phrase bereshith does not contain a definite article. The English translation “in the beginning,” does, of course, but no article is present in the Hebrew. The Hebrew article has the form ha-, though with the preposition b- the letter he would drop and we would have the fused form ba- “in the.” So the word reshith is “anarthous” i.e. it does not have an article.

Now the AP takes things a bit further, identifying reshith as not only anarthrous but in the construct state. Remember, all nouns in construct are anarthrous, though not all anarthrous nouns are construct, by any means.

Understand that the noun itself is identical in form whether it is construct or absolute. Nothing about it shows any marking for construct state. If it is not specifically identified as construct, by default is is absolute, as the SP takes it.

The lack of article is the sole indication for parsing reshith as construct. Is this enough? Hardly, in my opinion. Of course, other reasons are cited in favor of the AP, and we will touch on these below.

Identification from Context

What else can be an indication of construct state? What about context? In fact many nouns do not vary their form in construct state and are so identified only by relation to the word or words that follow. In the most typical instances (as we have described) a noun in construct state is followed immediately by a noun in the absolute state. These are very easy to identify.

E.g. melekh shalem “king of Salem” (Gen. 14:18)

We do see this with our word reshith elsewhere; in Genesis 10:10 it is unquestionably in the construct state:reshith mamlakto “the beginning of his kingdom.”

Now when I say “noun,” I mean this in a wide sense, proper and common, of course, as well as pronouns. Also nominal forms of verbs do occur, such as the construct infinitive. We do not have an OT example of one following reshith, but we do with another word meaning “beginning”:

bithillat shibtam

b-tehilla shebet-am

in beginning to-dwell their

“at the beginning of their dwelling there” (ESV)

“When they first lived there” (NIV)

We see a similar structure in Genesis 2:4

beyom ‘asot YHWH ‘elohim ‘erets weshamayim

b-yom ‘asot YHWH ‘elohim ‘erets we-shamayim

in day to-make YHWH God earth and-heaven

“in the day of the making of YHWH God earth and heaven”

“when YHWH God made earth and heaven”

In both examples the X noun is invariable, but is parsed as construct because of the Y forms that follow.

Finite Verb?

Do we have an analogous situation in Genesis 1:1, as proponents the AP maintain? What follows berishit is not any kind of noun or nominal, but a pure verb form, a finite verb. Note the difference with Gen. 2:4, where we have ‘asot, the construct infinitive “to make” or “making,” and not ‘asah, the finite verb form “he made.”

However Genesis 1:1 reads:

bereshit bara’ ‘elohim ‘et-hashshamayim we’et-ha’arets

b-reshit bara’ ‘elohim et ha-shamayim we-et ha-arets

in beginning created God OBJ the heaven and-OBJ the earth

Thus the AP parses reshith as the “X” of the construct chain with the Y being the finite verb bara’ with it’s subject‘elohim.

If this is the true syntax it is an extraordinarily strange thing. However, this is Hebrew, and extraordinarily strange things do sometimes occur. The most commonly adduced parallel is Hosea 1:2:

tehillat dibber YHWH

beginning spoke YHWH

This is so anomalous one could suppose the vowel pointing should read debar “word” instead of dibber “he spoke,” which would make this perfectly nominal (in both senses). But if the vowels are to be taken as written, we do have a finite verb as the “Y” element in a construct chain. Note however that tehillat (“beginning”)–unlike reshith–is readily identifiable as construct by form.

Another example is Job 29:2:

kime ‘eloah yishmereni

k-yeme ‘eloah yishmer-eni

“as days God watched me”

“as in the days when God watched over me”

Here also, the form unambiguously identifies the construct form of kime. A handful of similar structures may be identified (see GKC 422 §130.d), but they are always “extraordinarily strange.” And extraordinarily strange things happen rarely.

So is Genesis 1:1 another example? It is certainly imaginable, though not, I submit, the most likely syntactic understanding. By comparison, the SP is straightforward and ordinary.

Is the article expected?

The AP really only rises to a thinkable possiblity if with the SP we really do need the article in berishit but do not have it. In other words, in reading Gen. 1:1 are we really offput by the absence of the article? Ought we do expect it?

It’s no good of course, to reason back from the English translation. Just because we say “in the beginning” is no reason to expect an article in another language. The syntax of definite articles is frequently surprising, and particularly in prepositional phrases which have an adverbial meaning they may be absent.

A few English examples:

We employ some phrases with a distinction in meaning between articular and anarthrous versions:

e.g.: at school/at the school; in church/in the church; in the hospital (Am?)/in hospital (Br?)

Some are typically anarthrous:

e.g. at work, at home, at play

Some vary by dialect: an American might say “in the future I expect improvement,” whereas an Englishman might omit the article “in future I expect improvement.”

And we even say “from beginning to end” without articles on either noun.

Note that the LXX Greek translation of Gen. 1:1 has an anarthrous form in Greek: en arche. Now it may be supposed, and is frequently stated in the literature, that the Green is anarthrous due to the influence of the Hebrew. Perhaps this is the case, but note that ‘elohim is similarly anarthrous in Gen. 1:1 but the LXX translators rendered ho theos as articular.

John 1:1 similarly has the anarthrous en arche, though this is undoubtably an allusion to the LXX reading of Gen. 1:1. However, in NT Greek arche seems to be routinely anarthrous, particularly in prepositional phrases:

en arche (John 1:1,2Acts 11:15Phil. 4:15😉

ap’ arches “from (the) beginning (Mt. 19:4824:21Mark 10:613:19John 8:4415:27Acts 26:42 Pet. 3:41 John 1:12:71314243:811; 2 John 5,6)

ex arches “from (the) beginning” (John 16:4)

kat’ archas (Heb. 1:10) “at beginnings”

But what about Hebrew? What we would like to see is clearly anarthrous examples identical or analogous tobereshith and in the absolute state. The data are ambigous, as we do have several examples in Isaiah:

Isaiah 46:10 has the closest parallel with mereshith (“from the beginning”), anarthrous with the preposition min“from.” The example is weakened by the fact that this is prophetic literature, and poetic, which means the article is used very sparingly. We have the same problem with a synonym in Is. 40:2141:42648:16Jer. 17:12). Otherwise many, many examples where reshith is in construct because of following nouns or has possessive pronominal suffixes.

On the other hand, of the 51 instances of reshith in the Hebrew text, we have no example of *bareshith in the absolute with the article. In fact there is only one use of reshith with the article at all: Neh. 12:44 has lareshith“for the firstfruits.” Along with “contributions,” and “tithes,” here it refers to physical objects, a significant semantic distinction from what would be an adverbial use per the SP.  We also have Leviticus 2:12, the sole example ofreshith in absolute state, anarthrous, and not in a prepositional phrase, in prose, and it is likely simply indefinite. All others are in poetry such as Deut. 33:21 and Prov. 4:7.  Many, indeed most instances are unquestionably in construct state. Such is our data. We can therefore hardly claim to have any expectation in regard to the article or lack thereof in Gen. 1:1.

One other consideration which falls within textual evidence. The Masoretic pointing includes “accents” which include both disjunctive and conjunctive marks. This tells us what the scribes intended to be read as pauses or minor separation within verses and which words join together in a unit. If the AP holds, berishith links immediately with the following two words as a construct chain. It should be accented accordingly.

However, this is not what we have. The first word, bereshith, is marked with a major disjunctive accent, as shown in the figure below. It is that “backslash” marked m’yela in the illustration. The scribes, recording their centuries-old tradition, are showing us that berishith is not followed by anything in a construct chain. Again, the accents are not part of the original text, but neither is the shwa.

Bottom line here is whatever the shwa may seem to give in the AP’s favor, the accent pointing takes away.

The above graphic courtesy of John J. Parsons, Hebrew for Christians ( תודה רבה

So much for the primary sources. At this point the AP is imaginable, else no one would have imagined it. But to say that two options are possible is not to grant them equal probablity. Construct chains with finite verbs in the “Y” position exist, but as a handful of oddball examples within the thousands and thousands of construct chains in the Hebrew text. These are generally recognized as such because there is no other reading possible. In contrast, the SP simply has an anarthrous noun where we might have expected to have an article. The Hebrew text does not supply us with enough data to indicate how formidable an obstacle this actually is for the SP. At this point a native speaker competence would be very helpful.

We do have a stand-in for an available native speaker of ancient Hebrew. The standard parsing is standard because it is the way every ancient translation renders the text. By this measure, the lack of article was definitely no hindrance to the SP, nor did it lead any to propose the AP. The unanimity of this witness is highly significant. So what does the AP have in its favor?


We do have to admit a respectable pedigree in the person of the eminent 11th century AD French Rabbi known as “Rashi.” He proposed the AP in his commentary on Genesis, as proponents hardly fail to point out. He explain the meaning in this way:

At the beginning of the creation of heaven and earth, when the world was unformed and desolate, G-d said, “Let there be light.”

He explains his reasoning by saying that reshith is exclusively used in the construct state:

For the word reshith never appears in Scripture except when it is annexed to the following word. For example, “At the beginning of Yehoyakim’s reign,”[or] “The beginning of his reign,” [or] “The first of your corn crop.” Here, too, you must interpret “In the beginning El-him created” as if [it were written] “At the beginning of the creating.”

The logic is sound: IF the word is only and always used as the “X” element of a construct chain, it must be so here as well, and we have no option but to understand the AP.

With respect to the rabbi, however, the premise of his “IF” is manifestly false. It is perfectly true that what he describes is commonly the case, but not by any means without exception. By my count it is 82% or 41 of the 50 instances besides Gen. 1:1. The nine other instances are as follows:

  • Lev. 2:12 refers to the “firstfruits” but reshith is the “Y” element in the chain, therefore in absolute.
  • Deut. 33:21 refers to “the first part” but has no element following. In the absolute.
  • Neh. 12:44 mentions “firstfruits” in the absolute.
  • Prov. 4:7 refers to wisdom as the “main thing.” Again absolute state.
  • Is. 46:1 refers in general to the “beginning” as well as the “end,” both in the absolute.
  • Job. 8:742:12, and Hos. 9:8 have reshith in the construct, but with a possessive, e.g. “his beginning,” not with a “Y” element denoting the whole.
  • Ezek. 48:14 has a construct with a different relationship: “firstfruits of the land,” i.e. belonging to the land, i.e. the first part of things coming from the land, but not a part: whole relation with the land itself.

This being the case, I have to dispute Rashi’s conclusion, since his premise exaggerates the data.

Lexical Tendancy

A similar argument stated is that since reshith is “usually” in the construct state (and we note that for one reason or another, we can put this at 90% of the time, including with possessive pronouns. It is not sound reasoning, however, to suppose that Gen. 1:1 has a 90% probablity of being in construct state. Each instance needs to be evaluated in its own context. We have already seen that the text and context of Gen. 1:1 does not readily suggest a construct state.

Enuma Elish

This supposed parallel, AKA the Babylonian Genesis begins with a “when” clause. Its opening words, enuma elish, mean “when in the height,” and the following opening is said to show similarities to Genesis 1:

When in the height heaven was not named,
And the earth beneath did not yet bear a name,
And the primeval Apsu, who begat them,
And chaos, Tiamut, the mother of them both
Their waters were mingled together,
And no field was formed, no marsh was to be seen;
When of the gods none had been called into being,
And none bore a name, and no destinies were ordained;
Then were created the gods in the midst of heaven,
Lahmu and Lahamu were called into being…

The idea of citing this Babylonian text is the suggestion that a “when” clause is a standard opening for an ANE cosmogonic text, and that Genesis 1 could not fail to follow the rules. After all, proponents say, we see this pattern in the so-called second creation narrative in Genesis 2:

When no bush of the field was yet in the land and no small plant of the field had yet sprung up—for the LORD God had not caused it to rain on the land, and there was no man to work the ground, and a mist was going up from the land and was watering the whole face of the ground—then the LORD God formed the man of dust from the ground and breathed into his nostrils the breath of life, and the man became a living creature. (Genesis 2:5-7)

So then if one adopst the AP, Gen. 1:1 may be seen as falling into this pattern–so the argument runs. (1) When phrase, (2) list of negative conditions (3) state of preexisting matter, followed by (4) creative acts.

Now given the very nature of the subject matter, an account of the remote past in which that which did not exist was brought into existence, introductory time references and statements of the non-existence of certain items is not particularly surprising. Supposition of some kind of dependence or features of a common genre are rather forced in this argument. Whatever may be noted similarities between Genesis 1 and the Enuma Elish, the differences vastly outweigh them. It is fashionable to dwell on a purported connection between the Babylonian and the Hebrew texts, but it is far from clear that there is any connection beyond some superficial details. Nor is there any particular hint in Genesis that the author either was aware of the Enuma Elish or more to the point, thought his readers would be.

At any rate, it is entirely speculative, and scarcely overcomes the data of the text itself. Nevertheless, I have the impression that here is the most persuasive argument for the AP for some interpreters at least. Parallelism conquers all. It has a definite appeal to those with a desire to take the Scriptural text down a notch, which disposition is not difficult to observe in the world of Biblical studies.

Preexistent Matter

More of a consequence than a reason for adopting the AP, this yet provides a certain appeal. It is said that if the per the AP reading, matter is preesistent, or eternal, and ex nihilo creation is denied. This is not quite correct, as even so all the text would mean is that God’s initial step was the creation of amorphous matter, since it is the earth that is formless and void and what is beginning is the “creation of heaven and earth.”

Nevertheless, to a certain mindset, finding an option that finds in the Scripture a statement of eternal, preexistent matter is a consumation devoutly to be wished. It would not exactly be support for biological evolution, but this theory is something of an opportunistic agent, and any crack in the integument will do.

Scripture Interprets Scripture

This post has attempted to demonstrate that from a strictly textual and linguistic basis, the AP is non-impossible, but hardly rises high in probablity, the SP being without true difficulty and readily comprehensible–and not just potentially, but the actual construction understood by translators over the centuries.

More to the point, to those who take the Scriptural text seriously, its commentary on itself has a supreme value. We have already pointed out that all ancient translations, particularly the LXX follow the SP. So does the most direct NT allusion to it, John 1:1-2–a deliberate parallel:

In the beginning (en arche) was the Word, and the Word was with God, and the Word was God. He was in the beginning with God. All things were made through him, and without him was not any thing made that was made. (John 1:1-3)

Hebrews 1:10 refers back to it as well, though less directly through citation of Psalm 102:26:

And, “You, Lord, laid the foundation of the earth in the beginning (kat’ archas), and the heavens are the work of your hands; (Hebrews 1:10)

And Jesus’ own words serve as confirmation of the SP. “The beginning” is understood as the entire initial creation period in Genesis 1, not simply the initial formless condition as the AP would indicate:

He answered, “Have you not read that he who created them from the beginning (ap’ archas) made them male and female (Matthew 19:4)


Based on primary evidence, the text and immediate context, as well as lexical usage, the AP is an imaginable option to the SP, but does not have enough to commend itself to displace it. Ancient learned rabinnic endorsement of the AP is impressive, but equally emminent rabbis supported the SP, and all ancient translations follow it. Purported parallels to ANE literature holds a magnetic attraction to those with a certain mindset. But the analogia fidei and direct parallels of Scripture, to my mind at least, trump any supposed parallel with Enuma Elish and its ilk. In the end, I think it is most likely that the SP is correct and is what was intended by the original author.

No comments yet

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: