Editor’s Note: The following essay appears in the Spring 2020 issue of Eikon.
So much ink has been spilled debating and discussing the imago dei. Can anyone possibly improve our thinking on this topic? Is an attempt to do so arrogance?
Recent study undertaken on the primary sources since the publication of the second edition of Kingdom through Covenant in June of 2018 has led me to a better grasp and understanding of the ancient texts. I would like to focus here on the consensus in scholarship today and seek to show what can be improved or needs to be altered as far as this consensus is concerned. In the conclusion I shall seek to interpret what this means for human being and function in the world.
State of the Art
Let us first note a few important publications on the imago dei. Some show the consensus existing today while others represent the most recent treatments.
We begin with the treatment of dĕmut (likeness) and ṣelem (image) in the Theological Dictionary of the Old Testament (TDOT). The articles are by Preuss and Stendebach, respectively, and derive from the original German publications of 1974 and 1989, respectively. Both scholars assert that dĕmut and ṣelem are almost indistinguishable in meaning. Further, they assert that the prepositions bĕ (in) and kĕ (according to) are semantically indistinguishable and are to be understood in the sense of beth essentiae, or beth of identity. Stendebach concludes,
“in any event, v. 26b is not describing the content of humans being created in the divine image, since although 1:26, 28 do associate this notion with dominion over the non-human part of creation, 5:3 and 9:6 do not. Genesis 5:3 involves a genealogy in which Adam is said to have become the father of a son according to his image. Here the reference to dominion makes no sense. The same applies to 9:6, which justifies the sanctions against spilling human blood by recalling how God made humankind in his own image. Hence dominion over other creatures can only be a result or purpose of being made in the image of God.”
Articles by E. Jenni in The Theological Lexicon of the Old Testament derive from an original in German appearing in 1971, even though the English translation came out in 1997 after TDOT. His treatment agrees in essence with the results in TDOT.
The recent commentary of Jean L’Hour may be mentioned. This commentary, which appeared in 2016, is over 260 pages and deals only with Genesis 1–2:4a. It is the most detailed and extensive exegetical treatment in recent scholarship. The results of L’Hour’s study are in line with the authors of TDOT. In addition, L’Hour considers the Tell Fakhariyeh Inscription and concludes that dĕmut and ṣelem are indistinguishable in this ninth century BC Aramaic text. The newer Hebrew lexica, such as the Dictionary of Classical Hebrew (2010) and 18th Edition of Gesenius (2009), do not alter the picture significantly. Lastly, in a collection of essays from IVP in 2016, Catherine McDowell popularises her doctoral dissertation published in 2015 and adds material on Genesis 1. She considers dĕmut and ṣelem to be synonymous in both Genesis 1 and the Tell Fakhariyeh Inscription. As we will note later in this paper, her considerations of the divine image as sonship support the exposition given independently by myself, C. L. Crouch, and Gavin Ortlund in earlier publications. In summary, scholars have generally understood dĕmut and ṣelem to be virtually identical in meaning.
For the first and second editions of Kingdom through Covenant, I felt it sufficient to base my study on the description of the words dĕmut and ṣelem in Hebrew in the superb monograph of Randall Garr which appeared in 2003. While I continue to hold that the description of Garr is both accurate and even-handed, I learned interesting things from my own exhaustive analysis of these words carried out since the publication of the second edition of Kingdom through Covenant on June 30 of 2018.
Lexical Analysis of “Image” and “Likeness”
Lexical and semantic analysis is based primarily on three things: (1) context and usage, (2) cognate languages, and (3) ancient translations. Of these three approaches, usage is primary in establishing meaning.
First, ṣelem is found in seventeen instances in Hebrew and seventeen in Aramaic in the Old Testament. Setting aside the five occurrences in Genesis 1 and 5, six instances refer to images or statues of idols (Num. 33:52; 2 Kgs. 11:18 = 2 Chr. 23:17; Ezek. 7:20; 16:17; Amos 5:26). Three further instances occur in 1 Samuel 6:5, 11 when the Philistines captured the Ark of Yahweh and suffered from boils and mice. They made images of the boils and mice and put them in the ox-cart that carried the Ark back to Israel. Presumably, these images had an apotropaic value. One instance has to do with an image drawn or better etched (חקק) on a wall, possibly in a relief of some sort (Ezek. 23:14 Qr). Two occurrences in Psalms have to do with images that are phantoms or shadows (Ps. 39:7; 73:20), i.e. images that are abstract and non-concrete.
In biblical Aramaic, five instances of ṣelem refer to a statue Nebuchadnezzar saw in a dream (Dan. 2:31(2x), 32, 34,35), eleven refer to an idolatrous image or statue he built for his people to worship (Dan. 3:1, 2, 3(2×), 5, 7, 10, 12, 14, 15, 18), and one case has to do with the expression on his face which is described as “the image of his face” (Dan. 3:19, e.g. facial expression).Extensive usage in Akkadian reveals a similar pattern, referring to an image or statue of a god, king, or general image, to a figurine or bas-relief drawing, to a constellation or one’s bodily stature, and beyond that are metaphorical uses. Usage in Ugaritic, all epochs of Aramaic, and later phases of Hebrew are similar..The Septuagint usually renders ṣelem as εἰκών, although εἴδωλον is used in Numbers 33:52, ὁμοίωμα in 1 Samuel 6:5, and τύπος in Amos 5:26.
Turning our attention to the twenty-five instances of dĕmut in the Old Testament (not extant in biblical Aramaic, although the cognate verb דמה is found in Daniel 3:25 and 7:3), aside from three occurrences in Genesis, the bulk of the occurrences are in Ezekiel 1, 8, and 10, where Ezekiel is attempting to describe features in his visions. Sometimes he says x is like y, where the word dĕmut is used for “like” in English. Occasionally he employs the expression דְּמוּת כְ (Exek. 1:28). Daniel 10:16 כִּדְמוּת is similar. Rarely he speaks of דְּמוּת כְּמַרְאֵה־ (Ezek. 1:26, 8:2) or כְּמַרְאֵה דְּמוּת־ (Ezek. 10:1) or uses מַרְאֶה as a synonym.
The pair of instances in Isaiah (13:4; 40:18) function in a similar way to that of Psalm 58:5. They are abstract and non-concrete. In 2 Kings 16:10, Ahaz saw an altar in Syria and sent his priest in Jerusalem a sketch of the altar and detailed plans for construction (1984 NIV rendering is excellent). Our word is used in 2 Chronicles 4:3 to describe what looked like bulls below the rim in the casting of the bronze sea. Finally, Ezekiel 23:15 refers to an etching on a wall. This passage will be discussed shortly.
A cognate of dĕmut in Hebrew is the verb dāmâ and dēmot in Samaritan Hebrew. The related noun occurs throughout all phases of Aramaic, beginning with the Tell Fakhariyeh Inscription, to be discussed shortly. A rare occurrence in Akkadian means a ‘copy’ while in Arabic, a freeze-image or statue is signified. An indistinct figure or object is the meaning in Tigrinya, a derivative of ancient Ethiopic. The cognate most significant is Aramaic.
The rendering in the Septuagint is usually ὁμοίωμα (14×) or ὁμοίωσις (5×), εἰκών (Gen. 5:1), ἰδέα (Gen. 5:3), and ὅμοιος (Isa. 13:4).
Can we learn anything new from these data? Let us address directly the claim made often that dĕmut and ṣelem are synonyms or otherwise indistinguishable. First, we can observe from the cognate languages that, at first glance, Egypt and Mesopotamia have only one word for image. Conversely, Aramaic seems to be the only language besides Hebrew which offers both words in its vocabulary. The term ṣelem is a loanword in Arabic, and Wellhausen thought dĕmut was an Aramaic loanword in Hebrew. As we will see, in the bilingual inscription from Tell Fakhariyeh, the Akkadian part has only one word for image, whereas the Aramaic has two different words. Yet further research reveals that Akkadian employs the words tamšīlu and muššulu, derived from a root cognate to משׁל in Hebrew, in a way quite similar to dĕmut in Aramaic; and so Akkadian may have the distinction possible in Aramaic and Hebrew that I shall propose. In Akkadian, the word tamšīlu means (1) “likeness,” “effigy,” “replica;” (2) “image,” “resemblance,” “counterpart.” It can refer to statues, figurines in magic, buildings, or topographic features. The images can be concrete or non-concrete. The related muššulu can mean (1) “likeness” or (2) “mirror.” There may be more overlap in meaning between ṣalmu and tamšīlu in Akkadian than ṣelem and dĕmut in Aramaic and Hebrew, but a distinction is nonetheless possible, as we shall see.
Second let us observe that the ancient translators did not normally render dĕmut and ṣelem by the same terms in Greek or Latin. From this we can see they understood them as carrying a different nuance or meaning, however similar or synonymous they might be. They were not just stylistic variants for the ancient translators.
Thirdly, none of the major lexica or lexical studies observe that Ezekiel is the only biblical book besides Genesis which allows us to see both dĕmut and ṣelem employed side by side, nor do they make use of that text to determine whether or not the two words have a different nuance or are identical and completely interchangeable.
Is there a difference in the Old Testament between these two words? Yes, there is, I would claim. The term ṣelem normally refers to an image or statue of a god or human person. The emphasis is on how the image or statue represents this god or human person to the world. Conversely, the term dĕmut focuses on the concept of comparison and likeness. Unlike ṣelem in Hebrew or tamšīlu in Akkadian, dĕmut is never used in the Old Testament of a statue. Instead, the word focuses on the relationship of the copy to the original. Sometimes the word essentially functions precisely the same way as the prefixed preposition kaph.
While two words may be synonyms, what does this really mean in linguistic terms? Even when we are dealing with synonyms, we do not think that the field of meaning or usage of the two words is identical or overlaps perfectly. There is usually some slight difference in nuance between the two words.
Let us look at the usage in Ezekiel 23:14–15 where both terms occur together and also in the Tell Fakhariyeh Inscription from the ninth century BC where both terms are also found.
The text in Ezekiel 23:14–15 is as follows:
וַתּ֖וֹסֶף אֶל־תַּזְנוּתֶ֑יהָ וַתֵּ֗רֶא אַנְשֵׁי֙ מְחֻקֶּ֣ה עַל־הַקִּ֔יר צַלְמֵ֣י כַשְׂדִּ֔י֯ים חֲקֻקִ֖ים בַּשָּׁשַֽׁר׃
חֲגוֹרֵ֨י אֵז֜וֹר בְּמָתְנֵיהֶ֗ם סְרוּחֵ֤י טְבוּלִים֙ בְּרָ֣אשֵׁיהֶ֔ם מַרְאֵ֥ה שָׁלִשִׁ֖ים כֻּלָּ֑ם דְּמ֤וּת בְּנֵֽי־בָבֶל֙ כַּשְׂדִּ֔ים אֶ֖רֶץ מוֹלַדְתָּֽם׃
“But she carried her prostitution still further. She saw men portrayed on a wall, figures of Chaldeans portrayed in red, with belts around their waists and flowing turbans on their heads; all of them looked like Babylonian chariot officers, natives of Chaldea” (2011 NIV).
The word rendered “portrayed” by the NIV means “drawn,” “etched,” or “inscribed.” The term “figures” in v. 14 is the Hebrew ṣelem and the translation “all of them looked like Babylonian chariot officers” in v. 15 entails the word dĕmut. Contrary to authors of lexica and lexical studies, it seems easy to distinguish dĕmut and ṣelem in this text. The term dĕmut focuses on the relationship of the copy to the original. The term ṣelem, however, focuses on how the physical figures or images in bas-relief impacted those who saw them, i.e. the relationship of the copy to the larger world. The impact and power of the images is that they excited lust in the eyes of the beholder so that they sought political alliances with the Chaldeans. This is metaphorically pictured as fornication by Ezekiel.
Next, consider the Tell Fakhariyeh Inscription from the ninth century BC. Professor Alan Millard provides the English Translation in the well-known collection of texts called Context of Scripture. The term dĕmut occurs twice and the term ṣelem also occurs twice. In footnote 10 Millard states,
“[t]he monument is termed dmwt’ at two points and ṣlm at two others, both words clearly referring to the same stone figure. While remembering that Aram. and Heb. are not identical, this parallel use suggests no significant differences of meaning should be sought between the two cognate Heb. words used in a similar way in Gen. 1:26, 27; 5:3.”
This view is affirmed by the recent commentary of L’Hour and also in the lexical studies in TDOT.
Closer analysis may cause us to question this orthodoxy. I have subjected the Tell Fakhariyeh Inscription to a careful text-linguistic analysis as modelled by Aaron Schade. This is important for the literary structure of this text. Macrosyntactic signals clearly mark Focus and Topic and these changes in Focus and Topic correspond to divisions in the text vis à vis the literary structure. As Professor Millard himself recognises, the Tell Fakhariyeh Inscription (hereafter TF) actually represents two inscriptions. The first entails the dedication of an earlier statue; the second involves the rededication of the statue currently bearing the inscription.
The following literary structure is informed by all scholarly work on this text, but describes the text according to its content in simple terms:
Literary Structure of Tell Fakhariyeh Inscription
god Praise 2–6a
king Praise 6b–7a
king prayer 7b–10a
king imprecation 10c–12b
Identification/(king Praise) 13a
king prayer 13b–14
god Praise 15b–16a
king imprecation 16b–23
The attitude, emphasis, and tone in the rededication is different from that of the initial dedication. In the original dedication, Hadad-Yithi is king of Gozan alone; in the rededication he is king of Gozan, Sikkanu, and Azran. Obviously, he has prospered, and his kingdom has grown. In the original dedication he was a young king; now he is established in his kingdom and much more powerful. Notice in the original dedication, the majority of the text is devoted to praise of his god. The imprecation section occupies only a couple of lines. In the rededication, only two lines take up his relation to his god, and a full eight lines occupy the imprecation section. Randall Garr has noted this as well. Most importantly, what scholars have not noticed is that the term dmwt’ refers to the original statue in the first occurrence and to the relationship of the copy to the original in the second occurrence, while the term ṣlm refers to the second version of the statue in both its occurrences. So, both occurrences of dmwt’ focus on the relationship of the copy to the original and emphasise the vertical relationship of king to his god while the term ṣlm corresponds to the emphasis in the rededication section on the horizontal relationship of king to his subjects—the majesty and power of the king in relation to his world. Here I am simply adding to the detailed discussion of Randall Garr.
Admittedly the Akkadian version doesn’t draw out these distinctions, but the country ruled by Hadad-Yithi was originally Aramaic speaking and only secondarily a vassal of Assyria. So, the Aramaic text is primary vis à vis the Akkadian version. In the Akkadian translation, one instance of dmwt’ and both instances of ṣlm are rendered by ṣalmu in Akkadian; one instance of dmwt’ is not translated.
Pace Preuss, Stendebach, and Jenni in TDOT and THAT, L’Hour, Millard and others argue for no distinction between the terms. Randall Garr however, in a article in IEJ on the Tell Fakhariyeh Inscription and in his three hundred page monograph on the image of God, is more on target when he argues that dmwt’ emphasises the relation of the king to his god while the term ṣlm emphasises the relation of the king to his subjects. Notice that in line 15 of the Tell Fakhariyeh Inscription, the king states that he made the dmwt’ better than what it was before. In other words, the statue made for the rededication has a better likeness to the original than the first statue had. The sculpture was more realistic and the likeness more recognisable. And dmwt’ is used to express this rather than ṣlm.
Before turning to the instances in Genesis, let us consider the meaning of the prepositions bĕ and kĕ. Earlier we saw that the consensus in scholarship is that the prepositions bĕ and kĕ are semantically indistinguishable and are to be understood in the sense of beth essentiae.
In spite of the fact that the two prepositions are close in meaning, we must not assume that the meaning is identical. Randall Garr is correct to affirm that “the differential marking of each non-obligatory phrase suggests that each phrase has distinct meaning, at least in relation to one [an]other.” His careful and thorough linguistic analysis reveals that the preposition bĕ = “in” emphasises proximity, while the preposition kĕ = “as” or “according to” emphasises something similar, yet distinct and separate. Garr’s linguistic analysis is also supported by the exhaustive research of Ernst Jenni, who has produced an entire monograph on each of the three basic prepositions in Hebrew. One volume analyses all 15,570 instances of the preposition bĕ, a second all 3,000 instances of kĕ, and a third all 20,000 instances of the preposition lĕ (“to” or “for”) in the Hebrew Bible. Jenni concludes that, in fundamental meaning, kĕ stands between the opposition pair bĕ (marking an equating relation) and lĕ (marking a non-equating relation) as an expression of partial equation (and so also partial non-equation) of the semantic characteristics of two quantifications. Thus, again, bĕ indicates something locative and proximate while kĕ indicates something similar but distal and separate.
We have already seen that, although the words “image” and “likeness” share similar meanings, each has a different emphasis. In the Tell Fakhariyeh Inscription the word “likeness” focuses on the king as a suppliant and worshipper of his god and communicates sonship. The word “image” focuses on the majesty and power of the king in relation to his subjects.
These ancient Near Eastern data confirm and correspond to the use in the biblical text. The word “likeness” (דְּמוּת) in Genesis is closely associated with the creation of the human race, human genealogy, and sonship. It occurs in Genesis 1:26 in the creation of humans and again in 5:1, when this is recapitulated under the heading “Birth History of Humankind.” The third use is in 5:3 with the generation of Seth. The word “image” (צֶלֶם) is consistently used of man representing God in terms of royal rule. Putting the nouns and prepositions together, humans closely represent God in image, i.e., they actually represent his rule in the world. Humans are also similar to God in performing the action of creating human life, but not in the same way. Thus, bĕ emphasises a way in which humans are closely like God, kĕ a way in which humans are similar, but distinct. This interpretation also explains the reversal of the prepositions in Genesis 5:3. Here Seth shares precisely in the matter of generation and sonship but is only similar and not identical in the representation of his father’s image to the rest of the world.
We turn now to Genesis 5:1–3 where we have the term “likeness” twice with the preposition bĕ and, conversely, the term “image” once with the preposition kĕ.
“This is the book of the generations of humanity. When God created humanity, he made it in the likeness of God. Male and female he created them, and he blessed them and named them humanity when they were created. When Adam had lived 130 years, he fathered a son in his own likeness, after his image, and named him Seth.”
Catherine McDowell comments as follows:
“Seth is in some way similar to his father, yet he is not Adam, just as Adam and Eve are like God in some way, yet they are not God. The author gives no explanation of what constitutes the likeness, but the plain reading of the text suggests that Seth resembles his father simply because his father begat him. By analogy, humans correspond to God because God creates them. Thus, this correspondence is intrinsic to the relationship between Creator and created. When read in light of Genesis 1:26–27, to which Genesis 5:1–3 refers, the correspondence the author may have had in mind seems to be that of class. Seth is a human being, not a fish or a sheep, because his father is a human being. In short, to be created in Adam’s likeness and according to his image means that Seth was created according to Adam’s kind.”
Ten times prior to Genesis 1:26 we are told that grasses or fruit trees produce according to their kind or that God created creatures according to its kind/their kind. The implication is first that Seth belongs to Adam’s kind as a human being; and second, that some kind of kinship exists between humans and God.
As McDowell notes, the divine sonship of the king in the ancient Near East is an enormous topic. In addition to the examples discussed in Kingdom through Covenant to illustrate ṣalmu in Akkadian, she draws attention to passages I did not discuss, but which may possibly support the distinction I am making between “image” and “likeness.” Perhaps I may cite her illustrations at length so that nothing is taken out of context:
“Beginning with Tukulti-Ninurta I (1243–1207 BC) the divine-royal relationship in Assyria was expressed in terms of statue manufacture and divine birth. In the hymn from the Tukulti-Ninurta Epic, the king’s body is likened to “the flesh of the gods,” a phrase known elsewhere in the Assyrian Erra Myth as referring to the wood from which divine statues were made. He was “successfully engendered through/cast (ši-pi-ik-šu) into the channel of the womb of the gods” and, as a result, “He alone is the eternal image (ṣalmu) of Enlil,” whom “Enlil raised … like a natural father, after his first-born son.” The combination of birthing and manufacturing imagery is striking. Not only is Tukulti-Ninurta’s body likened to a divine statue, but the process of his creation is described both in terms of manufacture and procreation. Peter Machinist rightly concludes that here “image” identifies the physical body of the king with a divine statue. However, in this context, “image” may have been intended as a double entendre, referring to the king both as a “living statue” of the god and also as Enlil’s royal son. Although the hymn avoids explicit deification of the king, it certainly leaves the reader with the impression that Tukulti-Ninurta I, unlike any other human being, had a unique and special relationship—which finds its closest analogy in sonship—with the god Enlil.
“The opening lines of the Babylonian creation story Enuma Elish reinforce the idea that image and likeness terminology designated sonship. The account begins with the creation of the primordial gods: Apsu and Tiamat beget Lahmu, Lahamu, Anshar and Kishar. Anshar and Kishar then beget their firstborn son, Anu, who is described as the likeness (muššulu) of his father. The following line reads, “and Anu begot Nudimmud, his image.” Both examples define the father-son relationship in terms of image and likeness: Anu is the image (muššulu) of his father Anshar, and Ea is the “likeness, effigy, replica, image, resemblance, counterpart, or equivalent” (tamšilu) of his father, Anu. Although there are additional texts to which one could appeal, the biblical and extrabiblical examples just noted are sufficient to demonstrate that image and likeness language was indeed used in Mesopotamia to define the relationship of a god to a royal or divine son.”
Like Genesis 1:26, “image” is used of the body of the king as a statue of the god. There is a double entendre of divine kingship and sonship in the Tukulti-Ninurta Epic, but the emphasis is on kingship and so the term ṣalmu is used. Interestingly, in Enuma Elish the term ‘likeness’ (muššulu, tamšilu) is used when the focus is more on begetting and sonship rather than on royal status. This is a clear indication of the distinction I am proposing in a cognate language. So the examples adduced by McDowell not only support her general treatment, but also the finer point made in Kingdom through Covenant of the distinction between “image” and “likeness”: the word “likeness” emphasises the relation to the original and speaks of generation and sonship; the word “image” emphasises the representation of the original to the world and speaks of royal rule and status.
This discussion can be concluded by summarising evidence adduced in detail in Kingdom through Covenant. Nathan MacDonald has recently written on Genesis 1:26 as a text without a context. He points to the narratives in Genesis 2, Genesis 3, and Genesis 4 as further discussion and treatment of the meaning of the divine image. He also appeals to Irenaeus, who, in spite of bad exegesis at points, understood the important connection between protology and eschatology. This confirms the approach I have taken.
In Kingdom through Covenant, both in the First and Second Editions, I attempted to expound Genesis 1:26–28 in the light of Genesis 2–3, Genesis 5, Genesis 9, Psalm 8 and also Luke 3, Ephesians 4 and Colossians 3. Recent work by Catherine McDowell on Genesis 2–3, and also Gavin Ortlund and Richard Lints focusing on the later texts, have developed this further. McDowell’s argument for sonship is supported by Genesis 5, but Psalm 8 argues for the idea of kingship as related to the divine image.
Michael Jones, a ThM Student at SBTS, recently explored the notion of fictive kinship in covenants; we need to focus on this idea to flesh out all that is meant by “image” and “likeness” in Genesis 1:26–28. One aspect of covenant language in the Bible and in the ancient Near East is the use of family language to support the notion of covenant. In the Covenant at Sinai, the language in the covenant ratification ceremony in Exodus 24 clearly portrays Yahweh and Israel as “getting married.” In marriage, we have individuals who are not related by blood but who by virtue of the covenant of marriage are now more closely related than blood relatives. Marriage entails fictive kinship. Fictive kinship explains the communal meal eaten at a wedding and the communal meal on the mountain in Exodus 24. In the ancient Near East you don’t eat with humans who are not family. When individuals who are not related by blood get married, the first thing they want to do is to eat together to show that they are now closer than any blood/family ties. This is supported by the research of Scott Hahn in his large work Covenant by Kinship and also by the massive data collected by Paul Kalluveettil on terms employed in the ancient Near East and Old Testament for covenant where the exact term covenant is not used. There are many ways of speaking about covenant without using the word. So, many agreements and treaties borrow familial language. The Suzerain-Vassal treaties employ the language of father and son. This is to underscore the fact that in a covenant, parties have undertaken commitments and obligations as strong or stronger than family ties. Even the relationship of a king to his subjects is understood in these terms since one of the epithets for king in the ancient Near East is “father.”
Genesis 5 clearly features generation and sonship as characteristic of “likeness” and Genesis 1 and 2 features servant kingship as characteristic of “image.”
The final passage in Genesis mentioning the divine image is 9:6, where the basis for avenging a human life taken wantonly is the fact that we are made as the divine image. McDowell points out that later in the Torah, avenging blood is the duty of the “nearest relative” so that Genesis 9:6 affirms the connection between the divine image and kinship / sonship. In the Torah, damage to another person must be repaid as much as but not more than the damage caused (an eye for an eye, etc.) Damage to another person’s property must be repaid plus an additional amount (e.g. Exod 22:1). This difference in the law of retributive justice indicates the value placed upon human life as the divine image.
Implications for the meaning of “Image” and “Likeness”
Let us now consider the occurrences of dĕmut and ṣelem in Genesis in light of the lexical study. Then results from our earlier study can be added to this.
The one and only supreme creator deity announces to the divine council in Genesis 1:26 the decision to make ’ādām. The adverbial modifiers “in our image,” and “according to our likeness” indicate a vertical relationship between humans and God that can be described as obedient sonship and a horizontal relationship between humans and all creation that can be characterised as servant kingship. The preposition bĕ indicates that humans represent the creator God in the world precisely, while the preposition kĕ heading “likeness” shows that our generation is similar but not precise to that of God. As I explain in Kingdom through Covenant, God addresses the divine assembly but proceeds to create humans in his own image and likeness and disenfranchises the divine assembly by assigning the ruling function to the humans.
In reporting the execution of the divine decision, Genesis 1:27 employs only ṣelem with the preposition bĕ because this verse is preparing us for the role of humans in the world. Their royal status will result in representing God’s rule among creatures on the earth. Their binary sexuality will equip them to multiply as God planned.
The exposition in Kingdom through Covenant argues that ruling is the result of the divine image and not the image itself. It also demonstrates that the image applies to both male and female, since ’ādām is generic. Furthermore, since the grammar applies to the product and not to the process, the fact that humans are the divine image is not merely a description of their function and role but speaks of human ontology and structure as well. We are hard-wired for relationship with God and with all creatures.
The fictive kinship of “sonship” and the royal status of kingship force us to view these relationships as covenantal. This is crystal clear from the language used in the Bible and in the ancient Near East. Moreover, this exposition is both full-orbed, positive, and rich in describing the covenant relationship between humans and God and far surpasses the shallow so-called “covenant of works” described in covenant theology, as Richard Lukas has shown.
Peter J. Gentry is the Donald L. Williams Professor of Old Testament Interpretation at The Southern Baptist Theological Seminary in Louisville, Kentucky.
 Earlier expositions of the imago dei by Hoekema and Collins are moving in a direction that is helpful but lack the exegetical precision and sharpening of thought attempted here: C. John Collins, Genesis 1–4: A Linguistic, Literary, and Theological Commentary (Phillipsburg, NJ: P&R Publishing, 2007): 61–67 and Anthony A. Hoekema, Created in God’s Image (Grand Rapids: Eerdmans, 1986).
 H. D. Preuss, “דָּמָה dāmāh; דְּמוּת demûth,” TDOT, 3:250–260 (Grand Rapids: Eerdmans, 1978) and F. J. Stendebach, “צֶלֶם ṣelem image, model,” TDOT, 12:386–396 (Grand Rapids: Eerdmans, 2003).
 F. J. Stendebach, “צֶלֶם ṣelem image, model,” TDOT, 12:394.
 Ernst Jenni and Claus Westermann, Theological Lexicon of the Old Testament, trans. Mark E. Biddle, 3 vols. (Peabody, MA: Hendrickson, 1997), 339–42, 1080–85.
 Jean L’Hour, Genèse 1–2,4a Commentaire (Études Bibliques, Nouvelle série 71; Leuven: Peeters, 2016), 166–80.
 See also C. L. Crouch, “Genesis 1:26–27 As a Statement of Humanity’s Divine Parentage,” Journal of Theological Studies NS 61 (2010), 1–15 and Gavin Ortlund, “Image of God, Son of God: Genesis 5:3 and Luke 3:38 in Intercanonical Dialogue,” JETS 57/4 (2014): 673–88.
 W. Randall Garr, In His Own Image and Likeness: Humanity, Divinity, and Monotheism (Culture & History of the Ancient Near East 15; Leiden: Brill, 2003).
 AHw 1078 f.; CAD Ṣ 16:79–85.
 See DLU 783 and Comprehensive Aramaic Lexicon (http://cal.huc.edu/). Apparently ṣanam is an Aramaic loanword in Arabic, and the word is also attested in Old South Arabic and modern śḥeri. R. Meyer, J. Renz, and H. Donner, Wilhelm Gesenius Hebräisches und Aramaisches Handwörterbuch über das Alte Testament, 18. Auflage (Berlin: Springer, 2013), s.v.
 See E. Jenni, Die hebräischen Präpositionen, Band 2: Die Präposition Kaph (Stuttgart: Kolhammer, 1994), 57.
 See Comprehensive Aramaic Lexicon (http://cal.huc.edu/).
 AK. damtu = copy CAD D 3:74. See R. Meyer, J. Renz, and H. Donner, Gesenius Hebräisches und Aramaisches Handwörterbuch 18 (Berlin: Springer, 2013), s.v. דְּמוּת.
 H. Wehr, A Dictionary of Modern Written Arabic (Wiesbaden: Harrassowitz, 1979), s.v. duma.
 R. Meyer, J. Renz, and H. Donner, Gesenius Hebräisches und Aramaisches Handwörterbuch 18 (Berlin: Springer, 2013), s.v.
 CAD T 18:147–50.
 CAD M 2:281.
 Aaron Schade, “Fronted Word Order in Phoenician Inscriptions,” in Linguistic Studies in Phoenician in Memory of J. Brian Peckham, Robert Holmstedt and Aaron Schade, eds., (Winona Lake, IN: Eisenbrauns, 2013), 111–37.
 W. Randall Garr, “ ‘Image’ and ‘Likeness’ in the Inscription from Tell Fakhariyeh,” Israel Exploration Journal 50/3–4 (2000): 231–232 and idem, In His Own Image and Likeness: Humanity, Divinity, and Monotheism (Culture and History of the Ancient Near East 15; Leiden: Brill, 2003), 121–22.
 The concordance of Old and Imperial Aramaic Inscriptions by Schwiderski lists three instances of דמו and six instances of צלם excluding the occurrences in the Tell Fakhariyeh Inscription (see Dirk Schwiderski, Die alt- und reichsaramäischen Inschriften, Band I: Konkordanz [Fontes et Subsidia ad Bibliam pertinentes 1; Berlin: De Gruyter, 2004], 211 and 710). For דמו, the instance in Aḥiqar 201 speaks of “the colour of his face returning.” This is relationship of copy to original. In Kraeling 3, 20f, the signatories guarantee a house corresponding to the original. Finally, the Tell-Ḥalaf Altar speaks of the correspondence of the statue to the original. The instances of צלם in Daskyleion 1, Emar Stele Fragment, Nerab 1:3, 6, 12 and Nerab 2:2 are all in funerary monuments and focus on the relationship of the copy to the world. The image has power for others.
 See E. Jenni, Die hebräischen Präpositionen, Band 1: Die Präposition Beth (Stuttgart: Kolhammer, 1992), 84.
 Garr, In His Own Image and Likeness, 95.
 Ernst Jenni, Die hebräischen Präpositionen, Band 1: Die Präposition Beth (Stuttgart: Kolhammer, 1992), 11–40; idem, Die hebräischen Präpositionen, Band 2: Die Präposition Kaph (Stuttgart: Kolhammer, 1994), 11–12.
 A common error in the transmission of the text is to confuse bĕ and kĕ since the letters are so similar. Evidence from the Dead Sea Scrolls is not extant for 1:26, 27; 5:1, 3; 9:6. There is no clear evidence from the Septuagint, Syriac, Latin Vulgate, Targums, or early Jewish Revisors (Theodotion, Aquila, and Symmachus) for confusion of the prepositions. A few late medieval manuscripts show confusion only in 5:3, since the prepositions are reversed from 1:26, 27. Thus, remarkably, the textual transmission is solid in these texts.
 The Hebrew term tôlědôt is construed as a heading in the text.
 Translation by Catherine McDowell, “ ‘In the Image of God He Created Them’: How Genesis 1:26–27 Defines the Divine-Human Relationship and Why It Matters,” in Beth Felker Jones and Jeffrey W. Barbeau, eds., The Image of God in an Image Driven Age: Explorations in Theological Anthropology edited by (Downers Grove, IL: IVP, 2016), 35. See also her doctoral dissertation: Catherine L. McDowell, The “Image of God” in the Garden of Eden: the Creation of Humankind in Genesis 2:5–3:24 in Light of the mīs pî pit pî and wpt-r Rituals of Mesopotamia and Ancient Egypt (Shiphrut 15; Winona Lake, IN: Eisenbrauns, 2015).
 Catherine McDowell, “ ‘In the Image of God He Created Them’: How Genesis 1:26–27 Defines the Divine-Human Relationship and Why It Matters,” 35.
 Catherine McDowell, “‘In the Image of God He Created Them’: How Genesis 1:26–27 Defines the Divine-Human Relationship and Why It Matters,” 40–41. See Philippe Talon, The Standard Babylonian Creation Myth Enūma Eliš: Introduction, Cuneiform Text, Transliteration and Sign List with a Translation and Glossary in French (State Archives of Assyria Cuneiform Texts IV; Helsinki: Neo-Assyrian Text Corpus Project, 2005), 33.
 McDowell rightly sees human rule as a result of royal status (34). Failure to distinguish image and likeness, however, results in perceiving humanity’s identity as son of God, but not the covenant relationship between humanity and creation (35–42). McDowell’s discussion of Yahweh as father and humanity as son cites important texts in the Old Testament but does not put them into the metanarrative of Scripture as in Kingdom through Covenant. A number of the texts cited are descriptions of those given Adamic roles and therefore relate more to the point than McDowell thinks (39). From assessing the ancient Near Eastern data she concludes, “Humanity is defined both as God’s royal “son” and as living “statuettes” representing God and his rule in his macro-temple, the world. I have focused on the former because the connection between image and sonship has received far less attention in the commentaries and the secondary sources despite its fundamental importance for understanding what it means to be created in the image of God” (42). Thus, she admits focusing on sonship even though she acknowledges humanity royal status. For page references, see Catherine McDowell, “ ‘In the Image of God He Created Them’: How Genesis 1:26–27 Defines the Divine-Human Relationship and Why It Matters.”
 Nathan MacDonald, “A Text in Search of Context: The Imago Dei in the First Chapters of Genesis’, in D. Baer and R.P. Gordon (eds.), Leshon Limmudim: Essays in the Language and Literature of the Hebrew Bible in Honour of A.A. Macintosh (Library of Hebrew Bible/Old Testament Studies, 593; The Hebrew Bible and its Versions; London: T&T Clark International, 2013), 3–16.
 See chapter 6 in Peter J. Gentry and Stephen J. Wellum, Kingdom through Covenant (Wheaton, IL: Crossway, 2012, 2018).
 Gavin Ortlund, “Image of God, Son of God: Genesis 5:3 and Luke 3:38 in Intercanonical Dialogue,” JETS 57/4 (2014): 673–688 and Richard Lints, Identity and Idolatry: The Image of God and Its Inversion (Downers Grove, IL: IVP, 2015).
 Michael Jones, “Fictive Kinship and the Heart of Yahweh: How Adamic Sonship Entails a Loving Creation Covenant,” Paper Presented February 28, 2018 in 84100—OT Seminar, The Southern Baptist Theological Seminary.
 Scott W. Hahn, Kinship by Covenant: A Canonical Approach to the Fulfillment of God’s Saving Purposes (New Haven/London: Yale University Press, 2009) and Paul Kalluveettil, Declaration and Covenant (Analecta Biblica 88: Rome: Biblical Institute Press, 1982).
 M.-J. Seux, Épithètes Royales Akkadiennes et Sumériennes (Paris: LeTouzey et Ané, 1967), 33. Cf. CAD A 1:71.
 Cf. Psalm 9:12–13 (11–12 ESV) where Yahweh is the avenger of blood and also 2 Chronicles 24:25.
 Richard J. Lucas, “Re-examining Eden: The Creation Covenant in Theological Systems,” Paper Presented at the Annual Meeting of the Evangelical Theological Society, November, 2016. After completing the research for this paper, I came across a major work by Andreas Wagner. He is a specialist in ancient Near Eastern studies, and especially art forms. His research confirms in broad outlines the conclusions reached here as can be seen in the following citations from his work. First, Wagner shows how images in the Old Testament, whether physical or verbal must be understood: “The main thesis of this book is that the image of God’s body, as it is drawn verbally in the Old Testament, must be comprehended along the lines of the Ancient Oriental / Old Testament understanding of images. Pictures in our modern world refer to visible objects, they are understood as portrayals of real objects. In the Ancient Orient, pictures referred to objects in their ideal, typical conceived form, more or less independent of their visible aspect. This is combined with a corporeal concept which diverges from ours, in which the body always stands for the functions it exercises. Figures in human form in the Ancient Orient can, therefore, be understood to indicate functions of the body quite independent of visibility, without referring to the visibility of the parts of the body depicted at all. Consequently, verbal images of the body of God in the Old Testament, the anthropomorphic figure, can express the functions connected with core elements of the body without indicating a visual figure.” Andreas Wagner, God’s Body: The Anthropomorphic God in the Old Testament. Translated by Marion Salzmann. London: T & T Clark, 2019, 159. Second, here are Wagner’s conclusions about the image of God: “Humans are conceived to ‘represent.’ God’s cult image ṣælæm is conceived to express similarity, demût. Together they constitute (a merism) the whole person in his relationship with God. Mankind wields dominion vicariously for God on earth (cf. the mandate to rule), as God’s representative (B. Janowski), the mandatary (G. v. Rad). Communication between God and humans must work smoothly if the mandate is to be understood, and therefore similarity is a basic prerequisite. Secondly, humans must be able to act like God, less almighty and within the confines of human ability, but nonetheless capable of acting like God. Both these aspects, communication and the ability to act, lead us back to the similarity of the corporeal and of God and humans, as described previously. “ Ibid., 157. These conclusions are compatible with the conclusions here. Humans are created to have a covenant relationship with God on the one hand and the world on the other. The notions of obedient sonship and servant kingship define humanity both functionally and ontologically. Also, the priority of worship is determinative for implementing the mandate.
You, too, can help support the ministry of CBMW. We are a non-profit organization that is fully-funded by individual gifts and ministry partnerships. Your contribution will go directly toward the production of more gospel-centered, church-equipping resources.