BOUDIN, Chief Judge.

A federal statute makes it a criminal offense to "make or cause the telephone of another repeatedly or continuously to ring, with intent to harass any person at the called number." 47 U.S.C. § 223(a)(1)(D) (2000). James Tobin was convicted by a federal jury in New Hampshire of conspiracy to commit this offense, and of aiding and abetting another to do so, and now appeals. The events leading to the conviction are as follows.

In 2002, Tobin was New England Regional Director of the Republican National Committee. Prior to the November 2002 election, Tobin traveled to New Hampshire to coordinate VIP visits to the state. During the visit Tobin was approached by Charles McGee, Executive Director of the New Hampshire Republican State Committee. There ensued a conversation regarding a plan by McGee to disrupt the operations of the New Hampshire Democratic Party on election day.

During this conversation McGee asked for the name of someone who might be able to assist in a plan of this sort. Tobin provided the name of Allen Raymond, a longtime acquaintance, who owned a business that coordinated and designed telephone services for candidates and campaigns. Tobin and McGee did not speak again, but Tobin made a telephone call to Raymond to alert Raymond to expect McGee's call.

McGee and Raymond spoke together and e-mailed each other several more times and agreed upon the means of disruption — telemarketers would inundate specified numbers with hangup calls — and the price for it. McGee asked Raymond to check with legal counsel, and Raymond thereafter said that he had. None of these calls or any e-mails were made known to Tobin. McGee provided Raymond with six telephone numbers: five were for Democratic Party phones and one was for the firefighters union, which was offering rides to the polls.

Just as the polls were opening on election day, McGee's direct superior, John Dowd, ordered the operation called off. McGee then attempted to contact Raymond. However, for approximately 85 minutes, the phones at the targeted numbers rang almost continuously and the six telephone lines were blocked by repeated hang-up phone calls made by the firm that Raymond had earlier retained. Later Raymond called Tobin to talk about what happened, calling again after being contacted by police about the phone jamming.

On May 18, 2005, a federal grand jury returned a superseding indictment charging Tobin with crimes stemming from the phone tie-up in New Hampshire. The indictment charged violation of 18 U.S.C. § 241 (conspiracy to violate civil rights); conspiracy under 18 U.S.C. § 371 to violate 47 U.S.C. § 223(a)(1)(C), (D); and aiding and abetting these last two violations, 18 U.S.C. § 2(a).

McGee and Raymond each pled guilty to a violation of 47 U.S.C. § 223(a)(1)(C). McGee served seven months and Raymond's sentence was reduced to three months after his cooperation at Tobin's trial. Tobin proceeded to trial, which began on December 6, 2005. At trial the government's principal witnesses were McGee and Raymond. At the close of the government's case, Tobin moved for judgment of acquittal. The government then dismissed, with the district courts permission, all charges relating to section 223(a)(1)(C).

On December 15, 2005, the jury acquitted Tobin on the first count (conspiracy to violate the rights of voters) and found Tobin guilty of conspiracy to violate and of aiding and abetting a violation of section 223(a)(1)(D). Tobin moved unsuccessfully for judgment of acquittal, arrest of judgment and for a new trial. Tobin was sentenced on May 17, 2006, to 10 months' imprisonment, two years' supervised release, and a $10,000 fine.

Tobin's first and most far-reaching claim of error relates to the proper meaning of section 223(a)(1)(D)'s "intent to harass" requirement. From the outset, the district judge was concerned that the government was seeking to extend the statute from one directed at harassment of the called party to one embracing the disruption of telecommunications systems. In the end, the judge adopted a compromise, instructing the jury on the definition of "harass" as follows:

A person uses the telephone to harass another if he or she intentionally employs the phone in a way that is not meant as a good faith effort to communicate with a person at the number called and is done with an unjustifiable motive. So, for example, it is possible to intentionally cause the telephone of another to ring repeatedly and yet not violate the law if the caller is trying, in good faith, to contact someone at the number called.

If, however, the caller causes the telephone of another to ring repeatedly and the caller is doing so for reasons other than a good faith effort to communicate with someone at the number called, the law deems such conduct to be harassing. So, in the context of this case, the word harass means and describes conduct that is intentionally designed not to communicate, but instead to impede, distract, disrupt or undermine, in a substantial and not a trivial way, the ability of persons at the called numbers to communicate with others and to effectively go about their business.

On appeal, Tobin argues that "harass," in the present context, means to cause emotional distress in persons at the called number, that the jury should have been so advised, and that the "good faith" and "unjustifiable motive" language greatly broaden the statute beyond its permissible meaning. The government responds that the attack was not preserved in the district court and is also without merit.

It is true that Tobin did not ask the district judge to use the emotional distress language now urged. Understandably seeking the narrowest reading, Tobin asked the district court (in objection b. to the instructions) to define the term "harass" to mean "to threaten or frighten." Tobin's fall-back position on appeal is less restrictive, and wisely so, since nothing in the term "harass" limits it to threatening or frightening conduct.

This omission, unless the two versions are very close, arguably forfeits this claim — subject always to the plain error doctrine. Compare United States v. Fuchs, 467 F.3d 889, 900 (5th Cir.2006), with Arthur Andersen LLP v. United States, 544 U.S. 696, 708 n. 10, 125 S.Ct. 2129, 161 L.Ed.2d 1008 (2005). Whether the plain error test could be met need not be decided because we agree with a companion objection to the instruction which Tobin fully preserved, namely, that (quoting his objection f.):

The references to "an unjustifiable motive" and "reasons other than a good faith effort to communicate" dilute the intent requirement, which is a specific intent to harass, not just any unjustifiable motive or any reason other than a good faith effort to communicate.

We side with Tobin on this single issue. The district judge made a creditable effort to make sense of the perplexing statute. But in the end, the district court's "unjustifiable motive" and "good faith" language, used virtually to define "intent to harass," broadens the statute unduly. In considering the objection, we are required to consider just what Congress meant by harassment and the problem is not straightforward.

We begin with statutory language. New Hampshire Hemp Council, Inc. v. Marshall, 203 F.3d 1, 6 (1st Cir.2000). Black's refers to "[w]ords, conduct, or action (usu. repeated or persistent) that, being directed at a specific person, annoys, alarms, or causes substantial emotional distress in that person and serves no legitimate purpose." Black's Law Dictionary 733 (8th ed.2004). This core conception, echoed in other dictionaries, suggests that "emotional distress" would itself be too narrow a reading.

Another shade of usage, which the government points to in defending the instruction, is "harass" in a more tactical sense, as "to trouble by repeated attacks, incursions, etc., as in war or hostilities; harry; raid." The Random House Dictionary of the English Language, Unabridged 870 (2d ed.1987). But even here the word "trouble" invites attention to the likely reaction of the recipient — a concern borne out by the examples in the legislative history. E.g., 114 Cong. Rec. 4932, 4933 (1968) (statement of Rep. Staggers) ("Think . . . of the person called to the telephone countless times during the day and night only to have the caller hang up when he answers.").

The legislative history also suggests that a prime motive of the harassment provisions of the statute — primarily current subsections (C) and (D) — was harassing calls made to service personnel and their families during the Vietnam War. See, e.g., 114 Cong. Rec. 4932, 4933 (1968) (statement of Rep. Kornegay) ("[T]estimony from the Defense Department indicates that harassing telephone calls are now being used by a perverted few against the families of our servicemen who are serving overseas, particularly in Vietnam. These harassments have included everything from false reports of death or injury, to threats, demands for money for the Vietcong, and gloating comment on the actual death in combat of servicemen."). These were, of course, calls designed to make the recipients miserable — not to tie up the telephone.

Admittedly, the government offers two examples in the legislative history of a prior unenacted version of the statute — directed only at Washington D.C., but virtually identical to the version later introduced — that mention the difficulties caused by the tying up of the telephone. But both examples concern calls primarily of the ordinary harassment variety — one concerned a domestic dispute and the other was directed at an employee personally — even though they also disrupted business operations.

The district court's "bad faith-improper motive" instruction would include harassing conduct so defined but would also include almost anything else of which the jury might disapprove including conduct that was solely designed to interfere with telephone communications. We think that a Congress that sought to reach and outlaw attempts wrongfully to disrupt communications would have used quite different language ( e.g., "impede" access or use; "disrupt"), along the lines of state statutes that are expressly so aimed. It is also hard to think why a Congress seeking to protect access to unimpeded telephone communication would have been worried only about disruption caused by ringing (as opposed, for example, to cutting the line or sabotaging the gear where the drop cable enters the home).

Little circuit precedent exists one way or the other, but the case closest in point is mildly helpful to Tobin. In United States v. Bowker, 372 F.3d 365, 382-83 (6th Cir.2004), rev'd on other grounds, 543 U.S. 1182, 125 S.Ct. 1420, 161 L.Ed.2d 181 (2005), the Sixth Circuit construed subsection (C) (note 1, above), to read "harass" to have the same meaning as companion terms in the subsection ( e.g., "abuse," "threaten"). Presumptively, subsections (C) and (D) should be read together. Cohen v. de la Cruz, 523 U.S. 213, 220, 118 S.Ct. 1212, 140 L.Ed.2d 341 (1998).

In sum we think that to equate harassment with any repeat calling done in bad faith is to enlarge the scope of the statute. We read subsection (D) to require an intent to provoke adverse reactions in the called party and hold that a bad motive of some other kind standing alone is not enough. There might be good reason for a federal statute directed expressly to simple, deliberate disruption but that is a matter for Congress to determine.

On our reading, the instruction language was overbroad and clearly prejudicial to Tobin. The motive for the calls was unjustifiable, they were made in bad faith, and their effect was (in the words of the instruction) to prevent "in a substantial and not a trivial way the ability of persons at the called numbers to communicate with others." The government does not and could not make a harmless error argument so a remand is required. Neder v. United States, 527 U.S. 1, 19, 119 S.Ct. 1827, 144 L.Ed.2d 35 (1999).

Tobin says that instead of a new trial we ought to order an outright acquittal because the evidence was insufficient to permit a rational jury to convict on either of the two counts of conviction. United States v. Aponte-Suarez, 905 F.2d 483, 489 (1st Cir.1990). The two counts, conspiracy and aiding and abetting, both relate to the same underlying subsection (D) offense, but they differ in their elements, and we begin with the conspiracy charge.

The underlying offense requires that someone "make or cause the telephone of another repeatedly or continuously to ring, with intent to harass any person at the called number." The conspiracy offense occurs if the defendant agreed with another to commit the crime, whether or not the crime was actually committed. United States v. Feola, 420 U.S. 671, 694, 95 S.Ct. 1255, 43 L.Ed.2d 541 (1975); Iannelli v. United States, 420 U.S. 770, 777, 95 S.Ct. 1284, 43 L.Ed.2d 616 (1975). The general conspiracy statute requires as well an overt act, but such acts occurred in this case. United States v. LiCausi 167 F.3d 36, 46 (1st Cir.), cert. denied, 528 U.S. 827, 120 S.Ct. 79, 145 L.Ed.2d 67 (1999).

Tobin argues that if any agreement occurred to commit the offense, it was achieved "after all contact" between Tobin and the others had ceased and that Tobin never agreed to the disruption plan that forms the basis for the conspiracy charge. The strength of the argument is that prior to election day, Tobin had only one pertinent conversation with McGee and one with Raymond, and only after that did McGee and Raymond work out the concrete details.

But to comprise a conspiracy, a scheme need not be fleshed out. A drug distribution conspiracy exists once two persons agree to obtain cocaine and distribute it, even if the amount, timing, supplier, price and distribution method are all left open. United States v. O'Campo, 973 F.2d 1015, 1019 (1st Cir.1992). Nor need the agreement be explicit; the understanding that comprises the agreement can be inferred from the consolidation of efforts. Iannelli, 420 U.S. at 778 n. 10, 95 S.Ct. 1284.

There is a sound basis for this minimalist conception of agreement, Developments in the Law: Criminal Conspiracy, 72 Harv. L.Rev. 920, 933-34 (1959), and anyway the law is settled. Interstate Circuit v. United States, 306 U.S. 208, 226-27, 59 S.Ct. 467, 83 L.Ed. 610 (1939). For us, the issues are whether the evidence in this case permitted a rational jury to find beyond a reasonable doubt (1) that a plan to cause the jamming telephone calls was agreed to by Tobin and (2) that Tobin also shared the necessary intent as to harassment.

McGee testified that he first considered disrupting the Democratic Party's voter turnout efforts in response to a Democratic Party protest outside his office in October 2002. Vendors he contacted were unhelpful and three or four days before the election, McGee approached Tobin. McGee testified, "I don't know how much detail I went into with him, but I think I gave him the essence of the plan or the idea and asked him if he knew somebody who could help me."

The government then asked McGee, "And the essence of the idea was disrupting phone lines?" "Yes," McGee replied. "And making telephone calls to disrupt phone lines?" "Yes." "What doss Mr. Tobin say to you?" "He gives me the name of a gentleman by the name of Men Raymond and his phone; number." McGee then testified that Tobin said "something to the effect that he might be able to help you or he can help you or something like that," adding that if Tobin had told him "no," he would not have proceeded.

On cross-examination, the defense easily got McGee to repeat that he was unsure as to the level of detail arid to state that Tobin "did not agree to participate with [McGee] to violate the law." But the level of detail does not control so long as Tobin, with the necessary intent to harass, knowingly joined in a venture to cause the ringing of the phones; and no express agreement is required for conspiracy. Further, as a general rule, ignorance of the law is not a defense to a criminal prosecution. Cheek v. United States, 498 U.S. 192, 199, 111 S.Ct. 604, 112 L.Ed.2d 617 (1991).

Nevertheless, without Raymond's testimony, it is unclear whether a jury could have found that Tobin knew that the disruption — McGee's expressed goal — was to be caused by repeat calls, which would or might tie up the phone lines of the recipient. Raymond filled this gap. After explaining that his company was a telemarketing broker who hired others to do automated calls, voter identification calls or voter advocacy calls for election campaigns, he described his call from Tobin as follows:

Q: What does he say?

A: He says — he tells me that he'd like to talk to me about a phone project in New Hampshire and then explains the project to me as to what it would entail.

Q: And what does he tell you?

A: He tells me that it would entail jamming, essentially disrupting, democratic party and affiliated democratic organizations' efforts to Get-Out-The-Vote on Election Day.

Q: And after he says that, what, if anything, do you say next?

A: My response is that anything can be done. Yes, we can probably do it. I related to him an example of inadvertent phone jamming which involves sometimes when you have a large volume of calls going into one switch, sometimes you will inadvertently jam the switch, thus essentially jamming the phone lines. So my point there is that if it can be done inadvertently, it can be done intentionally, and then I tell him that we can do it, that anything is possible.

This call by Tobin was to introduce McGee to Raymond, and Tobin had no further contact with either until the plan had been (briefly) put into effect. But the testimony allowed a jury to conclude that Tobin knew that the disruption was to be caused by multiple phone calls made in order to tie up phone lines. The only remaining question as to agreement is whether a jury could rationally conclude that Tobin had so far associated himself with the emerging plan as to make him effectively a party to it.

This is surely a close call. The jamming mechanism described by Raymond was slightly different than the final version and Tobin never explicitly endorsed the plan. But nothing in the evidence or the cases cited to us by Tobin precludes a rational jury from concluding that Tobin did just enough to be a conspirator: that he knew the general means by which disruption would be achieved and agreed to it — indeed, implicitly endorsed it — by his introductory call and conversation with Raymond.

Tobin relies also on cases saying that "mere presence" at a conspiratorial event or a "mere introduction" of conspirators is not enough. But Tobin was not merely present, as in Andujar, 49 F.3d at 22; and in the "mere introduction" case, this court found no evidence that the defendant "in some sense promote[d] [the conspiracy], ma[d]e it his own, [had] a stake in the outcome." Aponte-Suarez, 905 F.2d at 491.

The more difficult question is posed by the intent requirements. To prove the voluntary participation required for conspiracy, the government must prove both the intent to agree (which we have just discussed) and the intent to commit the substantive offense. United States v. Ortiz, 447 F.3d 28, 32-33 (1st Cir.2006). Subsection (D) requires both an intent that the telephone be caused to ring repeatedly or continuously and an intent "to harass any person at the called number."

Tobin says that no evidence proves that he intended the repeated calls or telephone ringing. Probably he was indifferent to the method of jamming, but the intent required as to repeated ringing is merely knowledge that the use of repeated calls was contemplated, the repeated ringing being then foreseeable. Raymond testified that multiple calls were discussed as a method of jamming; effective jamming over a substantial period could foreseeably require persistent calling.

But an intent to harass was also required; repeat calls to a fire department to summon help would be lawful. Did an intent to harass exist if Tobin merely knew that anger and upset were almost certain to result from the carrying out of the scheme with its repeated ringing and blocking of communications? Or must Tobin have had a subjective purpose ( i.e., an aim or desire) to cause the subject to feel harassed? For the former, the evidence is sufficient; for the latter, arguably not.

Tobin's brief does not directly debate the legal issue, choosing instead — perhaps for legitimate tactical reasons — merely to assume that "purpose" is required. The government, adopting a broader definition of harassment than we accept, bypassed the problem now posed (if Tobin conspired, assuredly a purpose to jam could be inferred). There are good arguments on both sides; and the outcome may determine whether there will be a new trial or an end to the prosecution.

For centuries, common law judges, and then statute writers, have used the terms "intend" or "intent" without distinguishing between knowledge of consequences and a desire to achieve them. LaFave, Substantive Criminal Law § 5.2(a) (2d ed.2003); cf. Model Penal Code § 2.02 (1962). For obvious reasons, it is usually enough that (for example) the defendant knew that pulling the trigger would kill the person at whom the gun was aimed. See LaFave, supra, § 5.2(a).

Tobin says that subsection (D) defines a "specific intent" crime and so it does. But "specific intent" — although often equated with "purpose" — can as easily describe any additional intent requirement (whether knowledge or purpose) beyond the mental state — normally knowledge — required for the actus reus. LaFave, supra, § 5.2(e). Completed crimes that require "purpose," like treason, are comparatively rare.

Thus, an "intent to harass" could merely mean a knowledge that harassment — in the emotion-provoking sense — will inevitably occur from the forbidden act, even where, as may be the case here, Tobin was indifferent to the emotional effect. Perhaps a desire to harass is morally worse than knowledge that harassment will assuredly result; but the statute's language could be read to embrace both and one starts with a presumption that knowledge is enough.

Nevertheless, there are arguments for a different result here. The legislative history, already described, suggests that Congress' incentive to pass the statute was primarily to deal with calls having the self-evident purpose to harass. Further, making knowledge enough would leave the problem of calls that could cause foreseeable annoyance or upset and yet be justified by some legitimate aim ( e.g., to warn a sleeping resident of a fire). The need to interpolate a qualification could argue for a "purpose" reading.

The arguments we have mustered suggest that the issue may be close and should not be resolved without briefing. And, since in this court Tobin has asserted, but not developed, the "purpose" reading, we are hardly compelled to address the contention. Thus, the only basis for us to order an acquittal at this stage would be if a properly-instructed jury could not rationally find that Tobin knew that the recipients would feel harassed.

Obviously Tobin knew that the scheme if carried out would disrupt communication; but, as we have already noted, Tobin may not have cared one bit about the subjective reaction of the persons at the receiving end. So the question is whether a jury could think that Tobin "had to" foresee the consequent reactions (or be willfully blind to them, which is also enough, United States v. Rivera-Rodriguez, 318 F.3d 268, 272 (1st Cir.2003)), and to us the answer is yes.

Much less need be said about the adequacy of the evidence on the aiding and abetting count. That Tobin assisted in the substantive crime is patent; his call to Raymond was integral to the accomplishment of the scheme. The question again is of intent: "The state of mind required for conviction as an aider and abettor is the same state of mind as required for the principal offense." United States v. Valencia, 907 F.2d 671, 680 (7th Cir.1990).

If a jury could find that Tobin intended to harass, it could easily convict him of aiding and abetting; the crime does not have the agreement element that complicated matters for the government on the conspiracy count. The intent to harass element is far more debatable but what we have already said about that applies equally to the aiding and abetting count: if only knowledge of the effects is required, a jury could find the requisite intent.

In sum, we think a jury could convict Tobin under subsection (D) on the evidence presented at the trial unless a purpose to harass is required; the purpose issue, not developed by either side on appeal, is left for remand. We think it fair to add that despite the unattractive conduct, this statute is not a close fit for what Tobin did. if the government thinks this a recurring problem, it better seek an amendment.

The judgment of conviction and sentence is reversed; the case is remanded to the district court for further proceedings consistent with this opinion.

It is so ordered.