Or in Hebrew:
They were counted as the Lord had commanded Moses Those that were counted were as commanded, from the age of thirty years until the age of fifty years.
ואותן הפקודים היו במצוה מבן שלשים שנה ועד בן חמשים
Shadal cites Sefer haZikaron, which is a supercommentary on Rashi, that one needs to say כמצוה, thus changing the bet to a kaf:
(ברש"י צ"ל: כמצוה (ע' ס' הזיכרון.This would seem to be a minor edit, but perhaps it shifts the meaning.
What is "bothering" Rashi here? What is motivating him? The span of pesukim read:
One important point, I would suggest, is that the etnachta is on וְעַל-מַשָּׂאוֹ, so the word וּפְקֻדָיו is not joined with וְעַל-מַשָּׂאוֹ. How does one make sense of וּפְקֻדָיו אֲשֶׁר-צִוָּה ה אֶת-מֹשֶׁה? One important thing Rashi does is include the text from וּפְקֻדָיו until the end of the pasuk in his citation, thus noting that this was a statement by itself.
But that statement seems syntactically awkward. What is the function of the word asher? It almost reads like a subordinate clause -- אֲשֶׁר-צִוָּה ה אֶת-מֹשֶׁה. Thus, "his numbering, which the LORD commanded Moses." If so, it would be a sentence fragment. His numbering, which the Lord commanded Moses, was what?!
The answer is that the word asher is to be understood as kaasher. Thus, "and his numbering, was as the Lord commanded Moses." How so? As per the instructions at the beginning of pasuk 47 above, at the top of the instruction -- מִבֶּן שְׁלֹשִׁים שָׁנָה וָמַעְלָה, וְעַד בֶּן-חֲמִשִּׁים שָׁנָה.
Thus, to my reading, Rashi's concern is syntactic, and with making the text of the pasuk flow smoothly.
Who is the antecedent of וּפְקֻדָיו according to this reading? It doesn't really matter. It could be Moshe's counting. It could be the tribe of Levi -- "his" people who were counted. Or it could be his individual enumeration -- the visitation, or counting, of the individual Levite. However one wants to read it, the point is to smooth the reading of the verse. We can understand ואותן הפקודים as a translation of וּפְקֻדָיו, and thus וּפְקֻדָיו means approximately "his counted people." But we can cast it in other ways along this line of explanation as well.
Indeed, Ramban cites Rashi and understands the antecedent to be Moshe. Thus, his counting means Moshe's counting, and thus the rules of counting were "as the LORD commanded Moses." The alternative, Ramban notes, is along the lines of Ibn Ezra. And that is that the antecedent is the individual Levite. Thus, וּפְקֻדָיו means "his appointing." Because pakad means counting but can also mean assigning an individual role. Thus, that last pasuk has:
עַל-פִּי ה פָּקַד אוֹתָם, בְּיַד-מֹשֶׁה: Moshe appointed them based on Hashem's command
אִישׁ אִישׁ עַל-עֲבֹדָתוֹ, וְעַל-מַשָּׂאוֹ: each person to his job
וּפְקֻדָיו, אֲשֶׁר-צִוָּה ה אֶת-מֹשֶׁה: and each man's personal appointment was that which Hashem commanded Moshe.
So Rashi is not just motivated by a desire to smooth the reading, syntactically speaking, but to clarify the semantics of the pasuk. He holds that that etnachta divides וּפְקֻדָיו from וְעַל-מַשָּׂאוֹ, and then refers to Moshe's act of counting, as in the beginning of the pasuk עַל-פִּי ה פָּקַד אוֹתָם.
In this instance, I am not enamored with Siftei Chachamim's questions and answers.He writes the text pictured to the right. First, he notes the antecedent is Moshe (the pekudim of Moshe as translation of ufekudav), but then senses an awkwardness in the repetition of Moshe later -- it should have said אליו. He gives precedence in the Biblical style for this type of shift and repetition. As I noted above, I am not so convinced that the antecedent needs to be Moshe. Rather, one can say that read it as referring to the counting of each individual Levite, or the tribe of Levi. Perhaps this is due to the distinction of כמצוה vs. במצוה. If one says it was "in the commandment," the connotation may be that Moshe was here fulfilling the commandment, so Moshe should be the antecedent. But with כמצוה, each person's counting could be "in accordance with" that which Hashem commanded. Even though I don't agree with the premise, Siftei Chachamim's question here is a fair one.
The second question seems to miss the mark. He asks why Rashi sees the need to explain that it was מבן שלשים שנה ועד בן חמשים. He answers that there were specifications. In Behaalotecha, in Bemidbar 8:24:
other commands to count, such as in Behaalotecha, from the age of 25 and on. Thus Rashi is clarifying, and proving from the fact that אֲשֶׁר-צִוָּה ה אֶת-מֹשֶׁה refers to a previous command of Hashem, and thus must refer to the rules for counting specified a few verses earlier.
I sincerely doubt that this hava amina ever ran through Rashi's mind. The question was rather as cast by Ramban, as whether וּפְקֻדָיו referred to a counting (as per Rashi) or an appointment to a job (as per Ibn Ezra). By citing the specifications for the count from a few verses earlier, Rashi is clarifying that וּפְקֻדָיו refers to the count, and that אֲשֶׁר-צִוָּה ה אֶת-מֹשֶׁה refers to the specifications of the count.
Siftei Chachamim appears to miss the motivation of Rashi, and thus asks a baseless, under the extremely questionable assumption that Rashi is trying to teach us something new about the specifics of this counting. And thus he comes up with an extremely farfetched "problem" that Rashi is solving here. The command to count, was specified a pasuk or two above, with the rules for counting! Why would one even think that the rules for counting would be according to the specifications from Behaalotecha, rather that the specifications in the immediate context?!
This may be instructive to the whole enterprise of determining "what is bothering Rashi." For this is not the only time people misread the motivation, and then come up with a "problem" Rashi is solving that almost makes Rashi out to be a dope. Rashi was just making use of the words of the earlier pasuk to clarify his interpretation of that last phrase, and that it was referring to counting specifications.
The other explanation of Rashi, which Sifrei Chachamim cites from a godol echad, is also extremely weak. That parse is that וּפְקֻדָיו refers to those "appointed" over them, namely Eleazar and Itamar. And Rashi is innovating that these, too, must be within this age span. This is a fairly farfetched reading of Rashi, but it is motivated by the same misunderstanding -- that Rashi is coming to tell us new information about the specification, rather than coming to tell us that this is a counting specification.