Are Inaccurate Images of Christ Not Images of Christ?
Theological Letters
A quick note: I’m producing The East-West Series, a comprehensive 25-lecture exploration of the theological differences between Eastern and Western Christianity. We’re crowdfunding to complete production — watch the trailer and learn more at theeastwestseries.com.
“Dusty” wrote to me with an objection to icons that he heard in a Presbyterian Sunday School class. The minister insisted that icons are a violation of the second commandment. “Dusty,” having read my article on John of Damascus, offered replies in line with John’s case. The minister replied by saying that we don’t know what Christ looked like, so images of Christ are not in fact of Christ. He drew the comparison with carrying around a photo of a woman to whom I’m not married and telling people she’s my wife. “Dusty” thought on the reply and, days later, realized that this objection does not address the points he was raising (i.e., the Incarnation means the Son of God is now visible). Nonetheless, “Dusty” was curious what I would say to the objection, despite it missing the point. Below is my reply.
Dear “Dusty,”
Thanks for your email and for your kind remarks on my article about John of Damascus. I’m glad you found it helpful!
In answer to your question, I would offer just a few points. The first is that you are correct, the minister’s reply indeed sidesteps the question you asked. His strategy is what, in logic, is called a red herring, which is a logical fallacy. The term refers to an old practice of dragging a herring across a path when training hunting dogs. The test was to see if the dog could keep focused on the scent it was supposed to be tracking or whether it would get distracted by the smell of the fish. (Whether this practice is apocryphal or not, I don’t know, but that’s the supposed background of the term.) The original question you asked was whether an image of Christ violates the second commandment. The minister’s retort raises an entirely different question, namely, whether an image that is not of Christ but is viewed as such violates the commandment. These are two different issues.
I won’t rehash the case in favor of icons. Since you’ve read my piece on John of Damascus, you’ve read my overview of John’s defense of icons. Assuming I’ve read you correctly, the new challenge this minister is raising is this: We do not know what Christ looked like, so any image we make of him will not be of him but of some imaginary figure whom we call Christ; hence the image will still be idolatrous, since it will not in fact be an image of the Incarnate Son of God but of our imaginary Christ.
The key presumptions of this objection are two: (1) We do not know what Christ looked like, and (2) an image is an image of someone if and only if it accurately resembles them. Let’s start with the second point.
I find the claim to be an interesting one, primarily because it raises a host of fascinating issues about image-archetype relations, a topic that I think is underexplored but is rich with philosophical importance.1 Rather than beginning with representational images — that is, 2D or 3D attempts to capture the person’s likeness — I’d like to start with non-representational representations. To wit, names and symbols.
I want to start here for two reasons. The first is that, from what I can tell, the image-archetype connection that is central to the iconodule position also applies to names and symbols. Second, if I am correct on this, names and symbols provide an obvious counterpoint to presumption (2). For they demonstrate that the archetype-image connection need not be based on an accurate representation of the archetype’s appearance.
By way of review, you’ll recall from my article that John presumes a connection between the image and its archetype, and I discuss the background of this claim. In Plato, we see a discussion of the difference between non-substantial images (e.g., shadows), on the one hand, and substantial images (e.g., a reflection), on the other. When considering the nature of a substantial image, Plato notices that an essential feature of an image is that it is referential: It is the image of (name the archetype). Hence, there is an essential connection between an image and that which it images. Admittedly, the relationship is one sided; the image is dependent upon its archetype, not vice versa, but there is an essential connection between the two nonetheless. If memory serves, I also note in that article that this connection is recognized in the Old Testament itself. The exposition of the second commandment in Deuteronomy (here using the Septuagint, as the fathers did) does not use the word “idol” (eidōlo) but “likeness” (homoiōma). And the evident concern of the passage is that, because God is invisible and thus has no likeness, to worship the likeness of another is to fall into idolatry, offering worship not to your God but to the archetype of the likeness you make — be a celestial body, an animal, or a man. Hence, the notion of a connection between image and archetype is something that seems to be presumed in the OT prohibitions on idolatry.
Images, however, are not the only thing for which the OT presumes such a connection. Names seem to also display the same type of connection. That is to say, the very same connection between a likeness and the one who it is like seems to apply to names as well. A name, like a likeness, is essentially referential, being the name of something. Georges Berguer has made the case that, amongst ancient religions generally, names were seen as an image or replica of sorts, substantially distinct from that which they name but nonetheless connected to the archetype — very much in line with the image-archetype connection. We can see this theme throughout the Old Testament in both positive and negative ways. In the positive sense, if the name of God is invoked upon a country or person, then that place or person belongs to Yahweh; the name itself creates a connection with God, whose name is upon that person or place (e.g., Gen 48:16; Dt 28:10; Am 9:12). In the negative sense, we see this in the OT prohibitions on the names of the “gods.” The Israelites are strictly warned by God not to invoke the names of other gods, not even uttering their names (e.g., Ex 23:13).
Symbols provide yet another ancient example of a visual indicator of a being who is not represented according to likeness. We might think, for example, of the Ankh, for the goddess Isis, or the scarab, which symbolized not only recreation or resurrection but also the god Khepri. In neither case is the god depicted. Nonetheless, the symbol is an indicator of that god. And just as the OT prohibits the utterance of the names of the god, it also prohibits the use of such symbols in its prohibitions on talacements and magic (Lev 19:26; 20:27).
As an aside, I think it is worth noting that there is a growing body of literature on demonology in the Ancient Near East generally and the Old Testament specifically. In the light of such scholarship, it seems increasingly clear that the OT names a great many demons and tends to see the gods, not as imaginative figments of human invention, but as fallen angels. Hence, the concern about using their names, their symbols, and their images is because these names, symbols, and images are connected to real beings.
Lest I get into a tangent on demonology, which is always a tempting tangent for me, I’ll return to my main point. The minister’s presumption (2) seems to be problematic in the light of ancient names and symbols. For the presumption is that the connection between image and archetype is based on or contingent upon accurate visual representation. Hence, if the portrayal of Christ is inaccurate in its representation, then it is not an image of Christ. But the fact that names and symbols have an image-archetype connection and yet have no representation of likeness whatsoever seems to problematize this presumption.
Now, perhaps one could reply that once a visual representation of one’s appearance enters the equation, resemblance now matters. Perhaps images or likenesses, being representations of the person’s appearance, require accuracy to qualify as an image of that person. Let’s consider this possibility.
I’m not certain such a view would be so easily granted in the ancient world. We certainly have examples of sculptures, especially amongst the Greeks, which were evidently attempting to capture an accurate likeness of the individual they were imaging — or, at least, the skillfulness and precision of the sculptures we have would lead me to believe this. However, such efforts at accuracy were far from uniform across the ancient world or in ancient iconography. Think, for example, of ancient Egypt. Egyptian statues of Pharaohs were often reused, removing the name of the previous Pharaoh and inscribing the name of the new Pharaoh. When considering the rather basic relief sculptures of the ancient Pharaohs, we can see why. The physical representation was often (though not in all cases) extremely basic, looking generically human, with the defining characteristics being only the royal dress of the figure. The inscribed name was what one looked to for identification (or perhaps the event depicted), not the resemblance of the figure to the person it represented. In this way, such ancient icons sit somewhere between a representation in the contemporary realist sense and a symbol. Such icons certainly offer a representation of the person, but the representation is very basic — male, bearded, Egyptian, royal, and so forth. Hence, in these icons, we see an image of an individual but the designation as such is clearly not based on an accurate representation of the person’s physical features.
I find this precedent noteworthy because of its parallels with Byzantine iconography. Notice that ancient Christian iconography, especially in the Christian East, shows greater affinity for this symbolic representation than for either classical Greek sculpture or the later Renaissance realism of the Christian West. Many Byzantine icons look very similar in appearance, precisely because the representation of the individuals represented is very generic. Clearly, we have contemporary icons that aim at greater representational accuracy, and we see more realist traditions emerge in Russia and Romania, for example, but this is more of an exception than the norm in the history of iconography. Moreover, we see the ancient practice noted above (i.e., the naming of the individual depicted) in early Christian iconography. The name or initials of the Saint is written on the icon; the event depicted is named, and so on — a practice that continues to the present in Eastern Christianity. From what I can tell, in some cases, we really do not know what a given Saint looked like; we merely have an account of the life of that Saint, and from that we can gather some very generic points about gender, race, likely hair style or garments. Hence, the iconography itself is nearly symbolic in representation, with the written name indicating who is imaged. My point, in short, is that the idea that an image of someone must be an accurate representation of their physical characteristics to serve as an image seems contrary to these ancient sensibilities, both pagan and Christian.
I dare say that, while this ancient sensibility may seem alien to us, I am not sure that it really is. Consider a child’s drawing. We often see drawings of a family — a mom, a dad, a child — that has little representational accuracy. Perhaps it is nothing more than a set of stick figures. Or for the gifted child, we’re impressed that the people depicted have a body shape more closely approximating a human form. When shown the drawing and told that it is mom, dad, and Sally, we accept it as just that, an image of Sally with her mom and dad. In fact, many a drawing comes to mind that have these figures labeled, with “mom” written above mom, “dad” above dad, and so on. In many ways, these drawings are contemporary relatives of ancient images that offer very generic representations, accompanied by names to identify who is imaged. Are these drawings not of mom and dad? I’m inclined to think they are, in fact, drawings of mom and dad.
This consideration brings another point to mind. We tend to speak about good and bad representations. One might say, not to Sally, of course, but outside of her presence, that her representation of mom and dad is not very good. Here, I cannot help but think of Plato. Critical to Plato’s philosophy is the insight that we universally assess material objects as good or bad, well-formed or malformed. That object is a bad circle, for example. But bad in comparison with what? Plato’s theory of Forms is based, in part, on the recognition that we have in our mind an immaterial, ideal archetype against which we compare the material representation. The comparison itself is thus (essentially) relational: This is a bad circle (image) in comparison with the Idea of Circle (archetype). In a similar way, our talk of Sally’s (hypothetical) drawing as a bad drawing is an assessment of its representational accuracy. Her drawing looks nothing like mom or dad. But in saying this, we are still acknowledging a connection between the image and that which it images. It is indeed an image of mom and dad, but it is a bad image because of how divergent it is from its archetype. In order for this assessment to make sense, we must admit the connection between the image and its archetype. In other words, the connection is presumed despite the inaccuracy.
Here, I think we can see a tacit third presumption under the objection. To wit, it takes a binary approach to representation: The likeness is either accurate or it is no likeness at all. However, such an assumption is evidently false, since we speak of good and bad representations all the time, as in the case of Sally’s drawing. Poor representation does not undermine the fact that it is a representation nonetheless.
Before moving on to presumption (1), allow me to offer a word on the minister’s hypothetical: i.e., I carry around a photo of a woman who is not my wife and claim that it is my wife. The problem with this hypothetical is that it fails to accurately represent the objection. The religious equivalent of this is not that one has an icon of Christ that is a poor representation of his physical appearance. Instead, it would be that I have an icon of someone who is not Christ — say, the god Isis — and I say it is an image of Christ. Such a scenario would wade into the waters of idolatry. In fact, it seems to be the very concern expressed in Deuteronomy: i.e., because your God is invisible and has no likeness, any likeness you make will belong to someone or something that is not your God. But this is not the objection the minister raised. The objection was not that the icons of Christ are of a man or a deity other than Christ; the objection was that they were likely inaccurate representations. To modify the hypothetical to be apt, then, it would need to be something like this. I, as an artist, create a portrait of my wife. I carry a print of it in my wallet. When I show it to people, I tell them it’s a portrait of my wife, but I admit it’s not very good (i.e., it doesn’t look much like her). Would that mean the portrait is not of my wife? No. It is a portrait of her. It’s just not a very good one. Reframed in this way, we can see that the objection fails. The image is still an image of her, regardless of whether it is good or bad, well done or poorly done.
Having looked at presumption (2), and tacit presumption (3), let’s consider presumption (1). Is it true that our representations of Christ are inaccurate? I trust the foregoing shows that this question has little significance to the iconodule issue. But let’s consider it nonetheless.
I have no doubt that some icons are inaccurate, at the very least because the innumerable images of Christ throughout history vary considerably from one to another. But can we say with certainty that all are? According to tradition, and written record, we have plenty of images that date back to the time of Christ. Saint Luke, we are told, produced one of the earliest icons of Mary, the Theotokos, with the infant Christ and gifted the image to Theophilos, recipient of the gospel of Luke and the Acts of the Apostles. Eusebius of Caesarea, who was no fan of icons, complains of the vast number of images of Christ and the Saints that came down to the present day, indicating that the creation of images goes back to an even more ancient time. In fact, Eusebius goes into great detail describing a sculpture of Christ in Caesarea Philippi that was commissioned by the woman with an issue of blood who was healed by Christ. One further story worth noting is of the icon made without hands. So the story goes, Abgar, a ruler of Edessa in Syria, had leprosy and heard of a healer, Jesus of Nazareth. (On the plausibility of word reaching Edessa, consider Matt 4:24.) Abgar sent an artist to Christ to paint an image of him, but Christ instead sent the artist back with a cloth on which Christ made an impression of his face — hence, an icon made without hands. Abgar was healed in its presence, and Thaddeus was sent to baptize him, along with all in Edessa. Abgar placed the icon over the city gate, and there is an entire history of how it was lost, rediscovered, and lost again.
Such accounts as these give us reason to think that some of the earliest images of Christ were based on firsthand accounts — and in the case of the icon without hands, firsthand impressions. As argued above, the accuracy of the images of Christ seems immaterial to whether they are in fact icons of him. Nonetheless, the presumption that they are inaccurate seems to be baseless. The claim would seem to require either knowledge of what Christ did in fact look like, so that the images could be shown to be inaccurate by comparison, or certainty that no images were based on firsthand accounts. I presume the former is not being claimed, so the latter must be the working premise. But as the above accounts indicate, the premise is far from certain.
I hope that helps.
Sincerely,
Dr. Jacobs
Don’t forget to pre-order the East-West series at the lowest possible price. I sincerely appreciate the support in this effort.
For an example of the metaphysical value of image-archetype relations, see my article “The Metaphysical Idealism of the Eastern Church Fathers,” section IV.


The history of visual representations of Christ and the history of the Shroud of Turin share many remarkable and intriguing points of connection. See "The Hidden History of the Shroud of Turin" by Jack Markwardt for a detailed exploration.