concurring:
I join in the opinion of the Court and write separately to explain further why I believe the government did not satisfy its burden of proof as to the separate receipt of each of the four images in question. To be sure, internet crimes of this sort make for complicated prosecutions. Even familiar concepts like actus reus and mens rea can present new difficulties in this context, and analogies to our precedents in mail fraud or Mann Act cases will be imperfect at best.
There are many ways in which a defendant might receive four “visual depiction[s]” from the web. In the simplest example, he could visit a single page that contains four large images, all of which are automatically displayed on the screen and downloaded to his hard drive. In such a ease, we could debate whether the defendant had committed four separate acts of receiving an individual image, or the single *289act of receiving one web page. Our holding in Reedy strongly suggests that such a defendant has “bundled” his conduct by website, rather than by individual image, and should face only a single conviction. In a more complicated example, however, a defendant might visit a web page that has thumbnails of various images, and then individually select some of those thumbnails to enlarge them on screen. In that case, would we focus on the receipt of the first page, or the subsequent receipt of each image, or both? And what if the defendant views the same image twice during one visit? How many receipts is that?
These various scenarios — and one could easily imagine others — demonstrate that conceptualizing the “knowing receipt of any visual depiction” via the internet is much more complicated than it is in cases involving regular mail. Even with complete information about what the defendant did and what he knew at the time he did it, categorizing the criminal behavior is challenging. Our opinion in Reedy illustrates this difficulty well, and handles it with great skill. In this ease, however, we are not well situated to do the same.
As the majority correctly states, the government’s burden is to prove that Buchanan took more than one action to receive the four images at issue. Majority Op. at note 42 and accompanying text. After combing through the record, we are left with insufficient proof of exactly what Buchanan did to receive the four images for which he was charged, and as a result we are unable to classify the defendant’s conduct as one criminal act or four. We cannot be sure, for example, if the defendant visited a single website that had four large images on it, or if the images were selected and received individually.1 This is not to say that the former is necessarily a single criminal act, nor that the latter is necessarily more than one. It falls to future panels to decide these issues when they are adequately presented by the record. It is to say, however, that if the “ ‘unit of prosecution’ for a crime is the actus reus, the physical conduct of the defendant,” United States v. Reedy, 304 F.3d 358, 365 (5th Cir.2002), as it surely is, *290then we need to know more about the conduct of this particular defendant before we can sustain his four convictions.
We can be sure of the fact that Buchanan violated 18 U.S.C. § 2252. What we do not know is whether he violated it once, twice, or four times. Accordingly, I agree with the majority that it would be multi-plicitous to sustain four separate convictions in this case.
. The closest thing to a description of the defendant’s conduct with regard to the individual images comes from a question posed by defense counsel to one of the government’s witnesses. The exchange went as follows:
Q. So it’s fair to say that whoever is looking at the computer — and, again, we’re assuming it’s Mike. You didn’t actually see him at the computer, is that right?
A. That’s correct.
Q. Whoever is at the computer is obviously clicking on images and then clicking on other images, is that correct?
A. True.
Q. Fair statement?
A. Fair statement.
This provides us with some idea about how the defendant was selecting and receiving images, as do some of the exhibits introduced at trial, such as the list of downloaded files and their time stamps. However, this evidence hardly constitutes adequate proof that the defendant received the four images in question by way of four separate, knowledgeable actions. There were, after all, many legal and illegal images found on the defendant’s hard drive, and defense counsel's questions were not specific as to the four images at issue. It is entirely possible that those four images were received as a result of a single knowledgeable action on the part of the defendant rather than four. As for the time stamps, they are of little value. The government's own computer expert testified — and it is uncontro-verted — that the time stamps do not necessarily indicate when the pictures were received. Her exact words were, "you really shouldn't hang your hat on file dates and times.” Unfortunately for the government, they have done just that, because the record is devoid of any other indication of when or how Buchanan acquired the four images in question. Without more proof as to the specific images, I do not see how the four separate convictions can stand.