concurring in part and dissenting in part:
I join in parts I and II of the Per Curiam opinion, and I join as well in part IV as far as it goes; however, I respectfully dissent from part III. I agree with appellant that the trial court erred when it modified the standard instruction on robbery, which says that a defendant may be convicted of robbery if the jury finds that he used force to “take” the property of another, by telling the jury instead that it must find that force was used to “take or keep” the property. I would therefore reverse appellant’s convictions of armed robbery (count 1 of the indictment) and felony murder based on armed robbery (count 4). Since the instructional error affects only these two counts, I would affirm all the other convictions on the merits, subject only to a remand for partial resentencing as outlined in part IV of the Per Curiam opinion.
The underlying legal issue appears to be one of first impression in the District of Columbia, although it has arisen in the courts of several other jurisdictions. After reviewing the relevant decisions of those courts, I would elect to follow what I regard as the better-reasoned rule and accordingly would hold, in light of the facts of this case, that the modification of the standard instruction was reversible error.
The trial court initially read to the jury the standard instruction on robbery, CRIMINAL JURY INSTRUCTIONS FOR THE DISTRICT OF COLUMBIA, No. 4.46 (4th ed.1993).1 However, at the prosecutor’s request, the court agreed to *24re-instruct the jury with a modified version of the standard instruction. Defense counsel made a timely objection, but the court overruled it, saying, “It just seems common-sensical to me.... [I]f you say, ‘Let me take a look at your gun,’ and then turn around in the next second and steal the weapon, pointing the gun and threatening ... the taking and keeping it was against the will of the possessor of the property. Although, I guess, hypothetically, it could be a theft and ADW [assault with a dangerous weapon], in my view it could also be an armed robbery.” The court then gave a modified instruction which said in relevant part:
To take property means to get possession of it so as to be able to exercise control over it. The taking or keeping must be against the will of the complainant. ...
To establish a robbery, it is not sufficient that the defendant took or kept the property. He must have taken it or kept it by using force or violence.
[Emphasis added.]2
Appellant argues that the trial court committed reversible error by instructing the jury in this way. According to appellant, the amended instruction was an erroneous statement of the law that “imper-missibly restricted the jury’s consideration of relevant evidence, and wrongfully avoided the clear implication that the government had not [met] its burden of proof on this particular charge.” The government responds that the jury instruction did not offer a novel definition of robbery, but one that was consistent with the elements of that offense. To support this argument, the government cites cases from several courts that have “declined to impose strict temporal requirements concerning the exercise of force or violence in robbery cases.” The government also argues, in a footnote in its brief, that because this court has given a broad meaning to the phrase “immediate actual possession” in the robbery statute, see, e.g., Leak v. United States, 757 A.2d 739, 743 (D.C.2000), cert. denied, 534 U.S. 1054, 122 S.Ct. 644, 151 L.Ed.2d 562 (2001), this court should also adopt a broad construction of “take.”3
I
Our robbery statute, D.C.Code § 22-2901 (1996),4 states in relevant part: *25“Whoever by force or violence ... shall take from the person or immediate actual possession of another anything of value, is guilty of robbery .... ” The essential element of this type of robbery is the use of force, which is the primary distinction between robbery and larceny (or theft, as it is now known in this jurisdiction). See 67 AM. JUR. 2D Robbery § 3, at 50 (2003). There are, however, at least two views as to when this force must be applied to turn a larceny into a robbery. One such view is that force must precede or be contemporaneous with the taking, in the sense that the force must be used initially to gain physical possession of the property. Another view treats the occurrence as a continuing transaction, so that the use of force to retain the property or to escape may be sufficient to make the crime a robbery, even if the initial taking was not by force. See Kristine Cordier Karnezis, Annotation, Use of Force or Intimidation in Retaining Property or in Attempting to Escape, Rather Than in Taking Property, As Element of Robbery, 93 A.L.R.3D 643, 645 (1979). Courts are not in agreement on whether the use of force in retaining possession of property previously taken or in attempting to escape, rather than in the physical act of taking the property, supplies the element of force essential to make the offense a robbery. See 67 AM. JUR. 2D Robbei'y § 27. “Some courts ... have refused to uphold robbery convictions where force was applied subsequent to the physical act of ‘taking,’ ” while “[ojther courts, in contrast, have interpreted robbery as a continuous transaction that is not complete until the perpetrator reaches a place of temporary safety.” Ball v. State, 347 Md. 156, 185, 699 A.2d 1170, 1183-1184 (1997) (citations omitted), cert. denied, 522 U.S. 1082, 118 S.Ct. 866, 139 L.Ed.2d 763 (1998).
The states that subscribe to the latter view can be further divided into two subcategories. First, there are those with robbery statutes that expressly include the use of force to retain the property or escape with it. Second, there are those with statutes that do not explicitly include the use of force to retain the property or to escape, but in those states the courts have adopted a temporal view of the “taking” requirement that is so broad that force used in such a manner is still considered to have occurred during the taking. Because this sort of force is not expressly contemplated in our statute, the correctness of the trial court’s “take or keep” jury instruction hinges on whether this court adheres to the broad temporal view of the “taking” or the more narrow one which I shall discuss momentarily.5
The government urges us to adopt the position taken by the Maryland Court of Appeals in Ball v. State, supra. Prior to Ball, the Maryland Court of Special Appeals required that the force precede or be concurrent with the taking. See, e.g., Cooper v. State, 9 Md.App. 478, 480, 265 A.2d 569, 571 (1970) (“there must be evidence of actual violence preceding or accompanying the taking”); Giles v. State, 8 Md.App. *26721, 723, 261 A.2d 806, 808 (1970) (“To constitute robbery, the actual or constructive violence must precede or accompany the taking”). The Court of Special Appeals later abandoned this view in favor of the broader view that subsequent force in the face of resistance would constitute robbery. See Burko v. State, 19 Md.App. 645, 657-658, 313 A.2d 864, 871 (1974) (“when one commits a larceny and then displays a weapon so as to overcome the resistance of the witness, the crime is then elevated to robbery” (citations omitted)).
This issue was resolved in Ball. In Maryland, robbery is defined not by statute but by case law as “the felonious taking and carrying away of the personal property of another, from his person or in his presence, by violence or putting in fear, or, more succinctly, as larceny from the person, accompanied by violence or putting in fear.” Ball, 347 Md. at 184, 699 A.2d at 1183 (quoting West v. State, 312 Md. 197, 202, 539 A.2d 231, 233 (1988)). With this definition in mind, the court concluded as follows with respect to the timing of the force:
[T]he better view is that the use of force during the course of a larceny in order to take the property away from the custodian supplies the element of force necessary to sustain a robbery conviction .... If, as in the instant case, the use of force enables the accused to retain possession of the property in the face of immediate resistance from the victim, then the taking is properly considered a robbery.
Ball, 347 Md. at 188, 699 A.2d at 1185 (emphasis added). This holding makes clear that the Maryland courts adhere to the broader view of “taking,” and thus that in Maryland a taking is not necessarily complete once the defendant has gained physical possession of the stolen property.6
Michigan, on the other hand, provides an example of a jurisdiction that defines “taking” narrowly. Like the case law in Maryland and the statute in Colorado, supra note 6, Michigan’s robbery statute formerly stated, in relevant part: “Any person who shall, by force or violence, or by assault or putting in fear, feloniously rob, steal and take from the person of another ... any money or property” is guilty of robbery. MICH. COMP. LAWS § 750.530 (2003).7 In People v. Randolph, 466 Mich. 532, 648 N.W.2d 164 (2002), the Michigan Supreme Court overturned previous decisions8 that followed the “transactional approach” in which a completed larceny may be elevated to robbery if the defendant uses force after the taking and before reaching temporary safety, and “reasserted] that the force ... must occur before or contemporaneously with the felonious taking.” Id. at 551, 648 N.W.2d at 174. In Randolph the defendant stole merchandise from a department store and was seized by a security officer as he was leaving. The defendant broke free from the guard’s hold and swung at the guard in his effort to escape. Because of the assault on the guard, which occurred after the defendant gained possession and con*27trol of the property, the defendant was convicted of robbery rather than larceny.
After tracing the evolution of the transactional approach in Michigan case law, the court embarked on an exhaustive review of the common law of robbery to demonstrate how it was not in accord with the transactional approach used by the Michigan Court of Appeals. The court began by quoting Blackstone, who defined robbery as “the felonious and forcible taking, from the person of another, of goods or money to any value by violence or putting him in fear,” and declared that “[t]he taking must be by force, or a previous putting in fear .... ” Id. at 537-538, 648 N.W.2d at 167 (quoting 4 BLACKSTONE, COMMENTARIES at -241-242) (emphasis added by the Michigan court). As the court pointed out, a number of other commentators also agree with Blackstone’s view that, under the common law, the use of force must precede or be contemporaneous with the taking. For example, Professor Torcía has stated:
At common law, and in some states, force or threatened force ... amounts to robbery only if it is used to “take” property from the possession of another. Force or threatened force used thereafter, in order to retain possession of the property taken or to facilitate escape, does not qualify. At best, in such a case, the separate offenses of larceny and assault or larceny and battery are committed.
4 CHARLES E. TORCIA, WHARTON’S CRIMINAL LAW § 463, at 33-36 (15th ed.1996) (footnotes omitted).9 Consequently, the Michigan Supreme Court held that, at common law, a robbery required that the force be used before or contemporaneously with the larcenous taking, and therefore “the ‘transactional approach’ ... is without pedigree in our law and must be abandoned.” Randolph, 466 Mich. at 546, 648 N.W.2d at 172. Thus, unlike the Maryland court in Ball, the Michigan Supreme Court has adopted the narrow view of taking, which regards the taking as complete once physical possession is gained.10
II
In the case before us, it must be remembered that the exercise of force by appellant — namely, the shooting of Mr. Carpenter — did not occur until after appellant already had physical possession of the rifle.11 Given the divergence of the case *28law, I would conclude, with all due respect to the Maryland and Colorado courts, that the narrow view is the more appropriate one, and more consistent with the common law.12 I would hold accordingly that the trial court’s jury instruction was erroneous insofar as the word “take” in the robbery statute does not have such a broad meaning that it can include force used to effectuate the keeping of property. In the District of Columbia robbery retains its common law meaning, see United States v. Dixon, 152 U.S.App.D.C. 200, 203, 469 F.2d 940, 943 (1972), and the common law definition has only been expanded by statute to include stealthy takings. See Irby v. United States, 250 F.Supp. 983, 988 (D.D.C.1965), aff'd, 129 U.S.App.D.C. 17, 390 F.2d 432 (1967) (en banc). While there are, to be sure, differing views as to the temporal requirements of a “taking” at common law, see, e.g., Randolph, 466 Mich, at 554, 648 N.W.2d at 176 (Markman, J., dissenting),13 I believe that the rule adopted by the majority in Randolph is the more persuasive and authoritative one.
Moreover, well-settled principles of statutory interpretation favor the narrow view that the taking is complete once physical possession is gained. “It is a maxim of statutory construction that the language of a statute should be interpreted in accordance with its ordinary and usual sense, and with the meaning commonly attributed to it.” United States v. Thompson, 347 A.2d 581, 583 (D.C.1975) (citations and internal quotation marks omitted). A reliable method of discovering a word’s ordinary meaning is to look it up in the dictionary. In WEBSTER’S THIRD NEW INTERNATIONAL DICTIONARY at 2329 (1976), “take” is defined as “to get into ... one’s possession, power, or control.” This definition in no way suggests that a “taking” still continues after the item is in one’s possession and control. On the contrary, the phrase “to get into” plainly implies that it is the initial gaining of possession that falls within the definition, not any subsequent retaining of possession.
Reading a broad view of “taking” into our statute would ignore the fact that had the legislature intended for the statute to have such a meaning, it could simply have worded the statute differently. For example, Connecticut loosens the temporal requirement by stating that force need only be used “in the course of committing a *29larceny.” CONN. GEN. STAT. § 53a-133 (2001). Likewise, New York incorporates the phrase “in the course of committing a larceny” in its robbery statute. N.Y. PENAL LAW § 160 (Consol.2002). While not conclusive, the absence of this or similar language from the District of Columbia statute at least suggests a legislative intent to adopt the more narrow definition of “taking.”
The government points out nevertheless that in this jurisdiction the element of “force” in a robbery prosecution has long been held to encompass even the most minimal application of physical effort. Thus, for example, it was held long ago (1940) in the Spencer case, supra note 3, that “[t]he force used to remove the money from the pocket is sufficient to satisfy the statutory requirement of ‘force or violence.’ ” 73 App.D.C. at 99, 116 F.2d at 802 (citing Turner v. United States, 57 App.D.C. 39, 40, 16 F.2d 535, 536 (1926) (“the requirement for force is satisfied within the sense of the statute by an actual physical taking of the property from the person of another”)); accord, e.g., Ulmer v. United States, supra note 11, 649 A.2d at 298 (quoting Turner); United States v. Mathis, 295 U.S.App.D.C. 296, 306, 963 F.2d 399, 409 (1992) (“under D.C. law, a defendant may be convicted of a crime requiring proof of ‘force or violence’ when the only force used is that necessary to lift a wallet from a pocket”). Thus, under the government’s theory, the force that appellant used to pick up the rifle from Ms. Ashmon’s lap would satisfy the “force” requirement of the statute, and the error in the “take or keep” instruction would arguably be harmless.
The problem with this theory is that it was never made known to the jury. The government did not differentiate, either in its presentation of the evidence or in its closing argument, between the force necessary to pick up the rifle from Ms. Ash-mon’s lap and the separate act of force that occurred a few moments later when appellant shot Mr. Carpenter. Nor did the court’s instruction draw such a distinction; instead, it was focused, at the urging of the government, on the use of force to “take or keep” the rifle rather than simply to “take” the rifle. Given this record, I cannot find the instructional error harmless.14
To summarize, I would reject the broad view of “taking” and adopt the more narrow view, wherein the taking is complete as soon as the property at issue is in the possession and control of the robber. Thus a robbery does not occur if force is applied only after the initial taking, in order to effect an escape or to retain control. In such a situation, the offender may be deemed guilty of assault and larceny (or theft), but not robbery. Because the trial court instructed the jury otherwise, I would reverse appellant’s conviction of armed robbery, along with his felony murder conviction based on that same armed *30robbery. Since my colleagues do not agree, I respectfully dissent.
. Standard Instruction No. 4.46 states in part that to establish the elements of robbery, "the government must prove beyond a reasonable doubt ... that the defendant used force or violence to take the property ... [and] carried the property away ...." It states further:
To take property means to get possession of it, so as to be able to exercise control over it. The taking must be against the will of the complainant ....
To establish a robbery, it is not sufficient that the defendant simply took the property; *24[he] must have taken it by using force or violence.
[Emphasis added.]
.The evidence showed that the rifle (the property that appellant was charged with stealing) was voluntarily yielded to appellant by Ms. Ashmon, and that the shooting of Mr. Carpenter with that very same rifle took place a short but appreciable time after appellant acquired possession of the rifle, and at a location (outside the apartment) different from the place where the rifle was taken (inside the apartment).
It should be noted that appellant was charged with robbing Carpenter, and not Ash-mon, from whom appellant initially took possession of the rifle by picking it up from her lap. The only charge involving Ashmon was assault with a dangerous weapon, as to which appellant makes no claim of error other than the one which we have addressed (and rejected) in part II of the Per Curiam opinion.
. Appellant does not argue on appeal that there was insufficient evidence that the rifle was in Carpenter's immediate actual possession. See, e.g., Leak, 757 A.2d at 743 ("A bicycle lying two feet away from the owner is, undoubtedly, within the victim's immediate actual possession ... [when] the owner is aware of the attempted taking in a setting of force and violence”); Spencer v. United States, 73 App.D.C. 98, 99, 116 F.2d 801, 802 (1940) (removal of money from the pocket of trousers lying on a chair while the owner of the trousers was in bed with a prostitute a few feet away was a talcing from the owner’s immediate actual possession).
. Recodified as D.C.Code § 22-2801 (2001).
. This court has not yet decided which of these two approaches is correct. Appellant cites Lattimore v. United States, 684 A.2d 357 (D.C.1996), in which we held that the crime of robbery is comprised of larceny and assault, and that the taking and asportation requirements of larceny are satisfied once the defendant has acquired "complete and exclusive control over the stolen items.” id. at 360 (citing Groomes v. United States, 155 A.2d 73, 75-76 (D.C.1959)). We then said that "[the defendants’] complete and exclusive control [was] demonstrated as they physically held the items in their hands while restraining and threatening to shoot [the victim].” Id. Although Lattimore makes clear what is required for the taking element to be satisfied, it does not necessarily establish when the taking is complete. Lattimore is therefore of no assistance here.
.Colorado case law also adheres to the broad view of taking. See People v. Jones, 990 P.2d 1098 (Colo.1999), discussed in the Per Cu-riam opinion, ante at 22. Essentially as in Maryland, robbery is defined in Colorado as follows: “A person who knowingly takes anything of value from the person or presence of another by the use of force, threats, or intimidation commits robbery.” COLO. REV. STAT. § 18-4-301 (2002).
. The Michigan statute was completely rewritten in 2004 and no longer contains this language. See MICH. COMP. LAWS § 750.530 (2004).
. E.g., People v. LeFlore, 96 Mich.App. 557, 293 N.W.2d 628 (1980).
. See also 77 C.J.S. Robbery § 13, at 605 (1994) ("[t]he test is whether or not the taking of the property has been completed at the time the force or threat is used”).
. In Missouri the robbery statute defines "forcibly stealing” as the use of force for the purpose of "[preventing or overcoming resistance to the taking of the property or to the retention thereof immediately after the taking.” MO. REV. STAT. § 569.010(1) (1994). The Missouri Court of Appeals has recognized, however, that before the enactment of this statute, "[t]he rule in Missouri has traditionally been that the force ... necessary to prove robbery must precede or be contemporaneous with the taking of the property,” but the statutory language "extended the time at which the use of force ... may occur [so that it] can now occur after the taking .... ” State v. Kelly, 43 S.W.3d 343, 349 (Mo.App.2001) (citations omitted; emphasis added).
.It is of course well established in this jurisdiction that a dead person can be the victim of a robbery under our robbery statute. See, e.g., Carey v. United States, 111 U.S.App.D.C. 300, 304-305, 296 F.2d 422, 426-427 (1961). However, the cases that involve robbing a dead person are not applicable here, regardless of whether the intent to steal was formed after the victim was killed, because in those cases force was used to take property after the victim was either dead or unconscious, See, e.g., Ulmer v. United States, 649 A.2d 295, 298 (D.C.1994); Smothers v. United States, 403 A.2d 306, 313 n. 6 (D.C.1979). Here no force *28was used against Mr. Carpenter after he was shot.
. I note that Ball and the cases it relies on, such as People v. Estes, 147 Cal.App.3d 23, 28, 194 Cal.Rptr. 909, 912 (1983) (defendant used force to prevent a security guard from retaking property stolen from a department store), and People v. Tinsley, 176 Mich.App. 119, 120, 439 N.W.2d 313, 314 (1989) (victim and son gave chase, but abandoned their pursuit after the defendant drew a gun), are factually distinguishable from this case in one significant respect. In those cases force was used to overcome resistance by the victim in an attempt to reclaim the property, whereas in the case at bar there was no attempt on the part of anyone to reclaim the rifle from appellant once he had it in his hands.. Thus, even if I were inclined to follow the holding in Ball, it would not be dispositive on the facts of this case.
. See also 4 CHARLES E. TORCIA, WHARTON’S CRIMINAL LAW § 463, at 39-40:
Moreover, conceptually, the statutory extension is not necessarily inconsistent with the common-law theory of robbery[.] A thief who finds it necessary to use force or threatened force after a taking of property in order to retain possession may in legal contemplation be viewed as one who never had the requisite dominion and control of the property to qualify as a "possessor.” Hence, it may be reasoned, the thief has not "taken” possession of the property until his use of force or threatened force has effectively cut off any immediate resistance to his “possession.”
. I suspect that the issue that prompts the present discussion in both this opinion and the Per Curiam opinion is the result of careless drafting of the indictment. The indictment should have charged appellant with taking the rifle from the immediate actual possession of Ms. Ashmon, which is what the evidence actually showed; instead, however, it stated that the rifle was taken from the immediate actual possession of Mr. Carpenter. Cf. Joseph v. United States, 597 A.2d 14, 19 (D.C.1991). To be sure, the rifle was also — at least in theory — within Carpenter’s immediate actual possession as well as Ash-mon's, given such cases as Leak and Spencer; see note 3, supra. Nevertheless, if the indictment had been more carefully written, there would probably have been no occasion for the court to modify the standard instruction with the "take or keep” language, and this issue would never have arisen on appeal.