How Many Powers Created the World?

Anthony William · 13 August 2026 · 9 min read

Introduction

In the amoraic period, the rabbis’ answers to the “two powers” heresy grew in number and variety, and the list of Bible verses regarded as dangerous grew with them. Segal takes this as a sign that the heretical challenge was expanding, not fading. By then the term “two powers” covered apocalyptic, Christian and gnostic positions, and in Palestine the gnostic and Christian opponents dominated the scene. He cautions that all the rabbinic evidence must be weighed before anyone tries to identify the sectarians grouped under this one heading.

This article follows Alan Segal’s analysis in Two Powers in Heaven (Brill, 1977), pp. 121-134, of the traditions attached to R. Simlai and R. Yohanan. The question is what these traditions say, and what they allow us to conclude about the heresy they answer.

The Tanhuma Version

In Tanhuma on Leviticus 19:2 (Kedoshim 4), the verse Joshua 24:19, “he is a holy God,” gives the heretics an opening because the Hebrew for “holy” is plural. They tell R. Simlai that this shows two powers. He replies that the verse goes on in the singular, and that they would have a case only if it said “they are holy gods.”

Segal notes that the same exchange appears in many other places attached to other verses, and that this is the only version that names the doctrine “two powers.” It is also not the earliest version. He concludes that the term was not central at the start and was probably added in a late redaction, not earlier than the fourth century. Even here the text says only that the arguments resemble that heresy, which confirms that “two powers” had become a conventional label for a range of heresies whenever a verse could be read as implying plurality in God. The argument is confined to grammar, but nothing suggests the heretics argued only from grammar.

The Babylonian Talmud: Sanhedrin 38b

The fullest version is in Sanhedrin 38b. R. Yohanan says that wherever the minim find grounds for their heresy, the refutation is near. The plural verses and the nearby singulars are:

In most other versions the teaching belongs to R. Simlai, Yohanan’s student. Segal suggests either that Yohanan’s name entered because this section collects his anti-heretical sayings, or that Simlai learned the rule from him. The third-century date holds either way.

On method, Segal reads the pattern as follows: heretics pointed to a grammatical plural to show plurality in God, and the rabbi answered with a nearby singular. Some of these verses must reflect real disputes, and others were probably added by analogy. There is no evidence that any heretical argument took the form in which it is reported, and the heretics’ beliefs were no doubt more sophisticated. Daniel 7:9 shows this. The original argument there must have concerned the two manifestations, the “son of man” and the “Ancient of Days,” so the singular verb is a secondary layer of defence. The rabbis also avoided describing the heretical view in any detail, so as not to help spread it.

The Jerusalem Talmud: “How Many Gods Created the World?”

In the Jerusalem Talmud (Berakhot, cited by Segal as 12d-13a), the minim ask Simlai how many gods created the world. He sends them to the first man and Deuteronomy 4:32: the text says “he created,” not “they created.” He gives the same answer on Genesis 1:1, and then states his rule that the answer is close by wherever the minim go wrong.

They return with Genesis 1:26, Joshua 22:22, Psalm 50:1, Joshua 24:19 and Deuteronomy 4:7, and he answers each with a singular verb. Each time his disciples say he has driven them off “with a stick” and ask what he will tell them. To the disciples he says that man and woman cannot exist without each other or without the Shekhina (on Genesis 1:26), that the three divine names are one name like “Basileus, Caesar, Augustus” (on Joshua 22:22), and that they are like labourers, masons and architects (on Psalm 50:1). On Joshua 24:19 and Deuteronomy 4:7 others add that God is holy and near in every form, citing among others Exodus 15:11.

Segal draws these conclusions:

  • The principle matches R. Yohanan’s: set a nearby singular against the plural.
  • The rule suits Genesis 1:1 best, where the singular verb answers the question of how many deities were present at creation. The Palestinian rabbis made no further comment on that verse, so the argument may have started there and been extended.
  • At Genesis 1:26 the rule drives off the heretics but leaves the meaning of the verse unsettled. The rabbis gave separate explanations, which shows the rule was applied almost automatically.
  • Joshua 22:22 and Psalm 50:1 differ in kind, because they turn on the repeated names of God and not on a plural. Segal thinks they may be as old as the oldest “two powers” arguments. The example that follows them is out of place, which suggests they were inserted from another source.
  • The link between heresy and creation is explicit here, since the question is how many gods created the world. In Sanhedrin 38b the two themes are treated separately.

The versions send the heretics to different primeval authorities: the first days, the first man, the record of creation, or the Torah. All depend on Deuteronomy 4:32, and different emphases in the verse produce the different authorities. Segal adds that Deuteronomy 4:34 (“the Lord is God in heaven above and on earth below; there is no other”) makes the argument work against any dualism and not only against two creators, and could be aimed at heretical distinctions between the Lord and God.

The Verses One by One

Segal lists the verses regarded as dangerous in the third century and checks each against the rest of rabbinic tradition for further hints about the heresy.

  • Genesis 1:1: Already contested in tannaitic times, when Ishmael and Akiba disagreed about the particle et. In the third century the problem was the plural form of Elohim, and the same word occurs in Genesis 1:3 and Genesis 1:6. Segal infers that the defenders had a specific doctrine of plural deities at creation in mind, one used by many groups with Hellenistic Jewish traditions and not exclusively gnostic, apocalyptic or Christian. He also notes the rabbinic report that the Greek Bible reordered the verse, which Rashi explains as avoiding the “two powers” difficulty.
  • Genesis 1:26: Sectarian exegesis of this verse goes back to Philo and before. Simlai’s answer, that the next verse is singular, does not touch the likeness between God and man, who is being addressed, or the repetition of “Adam.” Segal reads the silence as deliberate: so many doctrines found support here that the refutation was meant to dismiss them all with as little description as possible.
  • Genesis 11:7: The same plural of majesty. Targum Onkelos is uncomfortable with God descending and renders it as “was revealed.”
  • Genesis 19:24: The repeated “from the Lord” is traced to R. Ishmael ben Yose. In another report R. Judah takes one of the two mentions to be the angel Gabriel. Segal says Gabriel was not treated as a separate power, but the reading shows that the divine name could be applied to a being other than the one God, which worried the rabbis. R. Eleazar’s rule that “and the Lord” means God and his heavenly court answers this. The verse also matters in some Christian exegesis.
  • Genesis 35:7 and 35:5: The plural for God is tied to a body of speculation about Jacob at Bethel, built on Genesis 29:10. Philo treats “place” as a name for God, as maqom became one in rabbinic usage, and discusses a “second god” there.
  • Joshua 24:19: Segal calls the verse puzzling and says opponents may have derived two gods from it, one just and one merciful, or one who creates and one who forgives.
  • 2 Samuel 7:23: The phrase “a nation and its God” troubled the rabbis. R. Eliezer takes it to mean idols, and R. Akiba takes it to mean the Shekhina, who is redeemed with Israel. So there is room for a divine manifestation other than God, as long as it stays within orthodoxy.
  • Daniel 7:9: Segal says far more speculation in heretical communities rested on these verses than the rabbinic comment suggests.
  • Psalm 50:1 and Joshua 22:22: Three names appear, and several scholars (Segal cites Marmorstein, Büchler, Bacher, Herford and Bergmann) conclude that Christian Trinitarian teaching was under attack. Segal notes that the Septuagint already rendered these verses to imply two powers, not three, and thinks the traditions served as a defence against more radical doctrines.

Conclusion

Segal’s conclusion is that the phrase “two powers” appears only in a late version, and that the traditions ascribed to Simlai form an anthology of verses against heresy, reduced to a plural and a singular. Behind that form are traces of teaching about a second principal figure in heaven and about the repetition of God’s name. Two passages may point to three powers, but whether the group was gnostic or Christian is hard to say, and the Greek text implies two. Only the creation verses, Genesis 1:1 and Genesis 1:26, clearly support more than two, and that reading is suggested by “many powers” traditions of creation, not required by the text.

He therefore holds that even in the amoraic period the basic distinction was between one power and two powers in heaven, dividing orthodoxy from groups the rabbis placed outside the community. “Many powers” was probably a later term for more complex hierarchies drawn from the creation account, as in some forms of apocalypticism, Christianity and gnosticism. Neither term should be assigned to a single group, since Jewish-Christian groups with elaborate intermediaries could be called “many powers,” and some gnostics held only two gods, the high god and the demiurge.

Sources and Further Reading

Primary Sources

Books

  • Alan F. Segal. Two Powers in Heaven: Early Rabbinic Reports about Christianity and Gnosticism. Leiden: Brill, 1977. pp. 121-134.

Published: 13 August 2026

Author: Anthony William

Tags: Rabbinic Literature, Two Powers, R. Simlai, R. Yohanan, Sanhedrin 38b, Talmud, Midrash, Alan Segal, Divine Plurality, Christian Apologetics