BASSETT, J.
Following a jury trial in Superior Court (Delker, J.), the defendant, Brian Craig, was convicted on one count of criminal threatening, RSA 631:4 (Supp. 2014); one count of witness tampering, RSA 641:5 (2007); and one count of stalking, RSA 633:3-a (Supp. 2014). The convictions were based on a series of messages that he posted on his Facebook profile page in April 2012 that were directed to the victim. At the conclusion of the State’s case, the defendant unsuccessfully moved to dismiss all three charges. On appeal, the defendant argues that the trial court erred in denying his motion to dismiss the witness tampering and stalking charges for insufficient evidence. He does not challenge his conviction for criminal threatening. We affirm.
I. Factual Background
The jury could have found the following facts. In late 2011, the defendant met the victim at a restaurant in Exeter where she worked as a bartender and waitress. The defendant initially came to the victim’s workplace with his brother or with friends. The victim interacted with the defendant only at work, and, according to her, their relationship consisted only of “very casual, very simple” customer-server communications. In time, the defendant began coming to the restaurant by himself, and the victim noticed that he stared at her. On one occasion, he came in alone, and told the victim that he came in just to see her.
In April 2012, the defendant mailed a letter to the victim at her workplace. The letter addressed the victim by name, and began: “So, you must’ve heard I was speaking highly of you on my Facebook page. I can tell, because you'are trying to hurt me.” Alarmed by the letter, the victim contacted the Exeter Police Department. Shortly thereafter, the victim received a second letter at work, in which the defendant stated that he “had to get a few things off of [his] chest” about their relationship before he could “say good bye properly.”
On April 22, Officer Chadwick of the Exeter Police Department served the defendant with a stalking warning letter. Chadwick explained to the defendant that the victim had complained about his behavior, and that the letter was a warning from the Exeter Police Department that “future stalking behavior” would result in prosecution for stalking under RSA chapter 633:3-a. Chadwick confirmed that the defendant understood the warning letter and the consequences of violating it. On the same day, the Exeter police served the defendant with a no-trespass notice from the victim’s employer, informing him that he was forbidden from entering the [365]*365victim’s 'workplace, and that if he did so, he could be arrested for criminal trespass. See RSA 635:2 (Supp. 2014).
The next day, the victim received a third letter at her workplace. The defendant wrote, “[I can] never give you another shot again, since you chose not to repair the damage you caused in having me banned from [the restaurant] for having spoken of it on the internet.” Although the victim had been told by the police that the defendant had mailed another letter prior to being served with the stalking warning letter, she was nevertheless distressed when she received it. The victim was so troubled that, later that day, she filed a petition for a temporary restraining order.
On April 24, the Superior Court (McHugh, J.) issued a temporary restraining order against the defendant under RSA chapter 173-B, which was served on the defendant the same day. See RSA ch. 173-B (2014) (protection of persons from domestic violence). The restraining order required the “[stoppage of the mail letters and no contact whatsoever, phone, email, et cetera.” The order also notified the defendant that a final hearing on the restraining order was scheduled for May 4, 2012.
Subsequent to service of the restraining order, the defendant continued to post statements directed to the victim on his public Facebook page. On April 27, the defendant posted:
Dear Kitty Kat:
I just wanted to remind you that since you would have to choose to look at the things I say to you on Facebook, that it means my butt is covered. Also, you are not allowed to do anything back to me all week, as it would constitute a breach in your end of the whole Restraining order thing. So technically, you are the one in cuffs. HA HA!
[Y]ou need to stop trying to beat me and start helping save people from death.
I think by the day in court you will have come around.
Now you see, [victim’s name], why it has to be you. Only you can wake up and say “Oh, there’s no beating him, I better help him or we’re all dead.”
[366]*366The next day, April 28, the defendant posted:
Dear Babe
[Y]ou are the one person I could never walk away from, unless I was made to. I am just asking you not to make me. . . . [Y]ou made it so I could not come back. You did so to see if I would care .... Well, damnit, I care! .... This is not goodbye.
Despite acknowledging that, “I know you want me to slow down a bit on here,” the defendant continued to post statements directed to the victim on his Facebook page:
So you want to push with this restraining order eh? Ha Ha, okay! Here’s what we will do. Since it won’t be resolved this Friday, and you intend to use my facebook posts against me, even though they are not a crime, I can retaliate with law too.... I can represent myself and beat you ....
Just tell the judge you are all set, and I will never speak your name again. Don’t forget to bring this post in with you.
You’ll have to lie under an oath of God to tell them you first became aware of my words on Facebook via my letter.
You don’t want to go to jail for perjury do you?
The document I have here does not mention my Facebook wall. You lose again[.]
HA HA. I mentioned Facebook in a letter, you mentioned your knowledge of it in your complaint, yet did not say not to talk about you on here.
[367]*367Later on April 28, the defendant posted four more messages, instructing the victim as to what he wanted her to do and say at the hearing scheduled for Friday, May 4, and threatening her if she did not comply:
[H]ere’s my proposal. On Friday, you can either tell the judge you are all set with me ... [o]r, you can drop all [the] charges and become an honest woman.
[G]oing to trial means the entire staff at [your workplace] gets put on the stand to answer the question “Did she view [the defendant’s] Facebook wall, prior to the letter in which he mentions it was received?[”] Since I know you have been viewing my wall for quite some time, I win.
Oh Schnookums! I forgot to mention... if you get me convicted of anything, I go to jail for a year, and everyone dies in the Apocalypse, and it will be all your fault. So, your options are to be all mine as of this Friday, or f**k off forever.
No, I want the order removed before Friday now. Or I will have you held accountable .... You go tell the judge that you were mistaken, and you’d like it removed____You’re a s**t! [S]o shut up and do as I say.
Well folks, I am going to go silent for the week, and let [the victim] eat s**t and rot in Hell.
[G]o tell them you were lying and you want to face the music for it.
You can tell the police the truth and drop the charges on [M]onday[.] No, right now, go there now. [I]f and when I receive documentation that you have dropped the charges we can start all over.. . .
[368]*368Several days after the court’s entry of the restraining order against the defendant, the victim, for the first time, decided to read the defendant’s Facebook page. She did so because the defendant’s first letter referenced his posts about her on Facebook, and because her mother, who had read the posts, warned the victim of “the extent and the severity” of the language in them.
Although the victim had a Facebook page at the time, she was not a “Facebook friend” of the defendant.1 However, because the defendant’s page was public, the victim found the defendant’s Facebook page simply by entering his name into the Facebook search tool. The defendant’s posts were contained in his Facebook “Notes,” which the victim could read by opening the “Notes” section of the defendant’s Facebook profile page.2
The victim spent “about three hours” reading the defendant’s posts about her. She was “appalled” and “scared” by the language he used in reference to her, and by reading her name in one of his posts. Consequently, she contacted the Exeter Police Department and reported the content of the Facebook page. In response, Officer Chadwick logged onto Facebook, found the defendant’s Facebook page, and read the multiple posts directed to the victim, many of which were written after the defendant had been served with the restraining order. On April 28, Chadwick went to the defendant’s home and confronted him with printed copies of the Facebook posts. The defendant admitted that he wrote them, but said that he was “expressing his feelings.” Chadwick then arrested the defendant.
A grand jury indicted the defendant for witness tampering, RSA 641:5; stalking, RSA 633:3-a, 1(c); and criminal threatening, RSA 631:4. A jury trial was held on December 11, 2012. At the conclusion of the State’s case, the defendant moved to dismiss each charge. The court denied the motion. The jury convicted the defendant on all three charges. This appeal followed.
On appeal, the defendant argues that the trial court erred in denying his motion to dismiss the witness tampering and stalking charges. He contends that the evidence was legally insufficient to convict him of the two charges.
[369]*369
II. Explanation of Facebook Technology Relevant to this Case
Facebook is a widely-used social media website, available for free to anyone with an e-mail account, whose stated mission is “to give people the power to share and make the world more open and connected.” Mazzone, Facebook’s Afterlife, 90 N.C. L. Rev. 1643,1646 (2012) (quotation omitted); see Democko, Comment, Social Media and the Rules on Authentication, 43 U. Tol. L. Rev. 367, 376 (2012) (discussing access to Facebook). Facebook and other social media sites are becoming the dominant mode of communicating directly with others, exceeding e-mail usage in 2009. Diss, Note, Whether You “Like” It or Not: The Inclusion of Social Media Evidence in Sexual Harassment Cases and How Courts Can Effectively Control It, 54 B.C. L. Rev. 1841,1842 (2013). With over one billion active users, Facebook is “revolutionizing the way people behave... and interact with one another in their everyday lives” through site functions that facilitate sharing information, such as a user’s “profile page,” the ability to send personal messages to other users, and by allowing users to become “Facebook friends” with other users. Democko, supra at 368, 375-76; see Ehling v. Monmouth-Ocean Hosp. Service Corp., 961 F. Supp. 2d 659, 662 (D.N.J. 2013).
A profile page “is a webpage that is intended to convey information about the user.” Ehling, 961 F. Supp. 2d at 662. “By default, Facebook [profile] pages are public.” Id. When a user shares something publicly, “anyone including people off of Facebook can see it.” FACEBOOK, http://facebook.com/ help/211513702214269?refid=69 (last visited Jan. 22, 2015); see also Diss, supra at 1844 n.17 (“Public information is available to anyone, even to people without an account on [Facebook].”). Alternatively, Facebook users can restrict access to their Facebook content using Facebook’s customizable privacy settings. Ehling, 961 F. Supp. 2d at 662. “Access can be limited to the user’s Facebook friends, to particular groups or individuals, or to just the user.” Id.
III. Analysis
To prevail on a challenge to the sufficiency of the evidence, a defendant must show that no rational trier of fact “could have found the essential elements of the crime beyond a reasonable doubt, considering all the evidence and all reasonable inferences therefrom in the light most favorable to the State.” State v. Germain, 165 N.H. 350, 354-55 (2013) (quotation omitted). We examine each evidentiary item in the context of all the evidence, not in isolation. Id. at 355. Circumstantial evidence may be sufficient to support a finding of guilty beyond a reasonable doubt. Id. Further, the trier of fact may draw reasonable inferences from facts proved [370]*370and also inferences from facts found as a result of other inferences, provided they can be reasonably drawn therefrom. Id. When the evidence as to one or more elements of the charged offense is solely circumstantial, the defendant must establish that the evidence does not exclude all reasonable conclusions except guilt. Id. at 361. Because a challenge to the sufficiency of the evidence raises a claim of legal error, our standard of review is de novo. State v. Kay, 162 N.H. 237, 243 (2011).
A. Stalking
We first address whether the State presented sufficient evidence on the charge of stalking. See RSA 633:3-a, 1(c). The defendant’s arguments challenging the sufficiency of the evidence are closely intertwined with questions of statutory interpretation. For example, the defendant contends that there was insufficient evidence that he stalked the victim because he did not take an “action to communicate” with the victim as required by the definition of “contact” in RSA 173-B:1, IV. Therefore, determining whether there is sufficient evidence to convict the defendant of stalking requires us to interpret the stalking statute, RSA 633:3-a, 1(c), and the definition of contact in RSA 173-B:1, IV.
We begin our analysis by outlining the pertinent portions of the statutory scheme. In order to convict the defendant of stalking in violation of RSA 633:3-a, 1(c), the State had to prove that the defendant, after being served with a protective order issued pursuant to RSA chapter 173-B that prohibited contact with the victim, “purposely, knowingly, or recklessly engage[d] in a single act of conduct that both violates the provisions of the order and is listed in paragraph 11(a).” RSA 633:3-a, 1(c) (emphases added); see RSA 633:3-a, 11(a). Here, the State charged that the defendant: (1) engaged in an “act of communication, as defined in RSA 644:4, II,” see RSA 633:3-a, 11(a)(7); and (2) that this act of communication violated the provision of the restraining order that required “no contact whatsoever, phone, email, et cetera.” RSA 644:4, II defines “communicates,” in relevant part, as “imparting] a message by any method of transmission, including .. . electronic transmission.” RSA 644:4, II (2007). RSA 173-B:1, IV defines “contact” as “any action to communicate with another either directly or indirectly, including, but not limited to, using any form of electronic communication, leaving items, or causing another to communicate in such fashion.”
The interpretation and application of statutes present questions of law, which we review de novo. See Deyeso v. Cavadi, 165 N.H. 76, 79 (2013). “We are the final arbiters of the legislature’s intent as expressed in the words of the statute considered as a whole.” State v. Dor, 165 N.H. 198, 200 (2013). “When interpreting a statute, we first look to the language of the statute [371]*371itself, and, if possible, construe that language according to its plain and ordinary meaning.” Id. “We do not read words or phrases in isolation, but in the context of the entire statutory scheme.” Id. “Our goal is to apply statutes in light of the legislature’s intent in enacting them, and in light of the policy sought to be advanced by the entire statutory scheme.” Id.
The defendant does not dispute that he was served with the domestic violence restraining order on April 24, or that he subsequently posted the statements at issue on his Facebook page. Thus, the State had to prove that the defendant, by posting on his own public Facebook page after he had received the restraining order, engaged in a single act of conduct that constitutes: (1) an “act of communication”; and (2) “contact” pursuant to RSA 173-B:1, IV that violates the April 24 restraining order.
1. Act of Communication Pursuant to RSA 6W4, II
The trial court concluded that “the nature of what [was] written and how [it was] written . . . suggested] that [the defendant’s posts were] a communication directed at [the victim] in a public forum,” and that “the definition of communication [in RSA 644:4, II] is broad enough, certainly, to cover . . . these posts on Facebook.” Although the defendant asserts on appeal that the trial court erred in its interpretation of RSA 644:4, II, he makes only a passing reference in his brief to the issue; therefore, he has failed to develop this argument sufficiently for our review. State v. Young, 159 N.H. 332, 337 (2009).
2. Contact Pursuant to RSA 173-B:1, TV
The defendant’s argument that he did not “contact” the victim in violation of the restraining order has two main components. First, he asserts that his conduct is insufficient, standing alone, to constitute an “action to communicate” as required by RSA 173-B:1, IV. Second, he argues that the victim’s affirmative act of searching for and reading his Facebook posts precludes his conduct from constituting “contact” as defined above.
We first consider the defendant’s argument regarding his own conduct. The defendant contends that, in order for his conduct to constitute “contact” pursuant to RSA 173-B:1, IV, he must “be the actor not only in the creation of the message, but in the conveyance of it to the protected person.” The defendant argues that his Facebook posts cannot constitute contact because he merely posted publicly online without sending the posts directly to the victim, and, therefore, did not take an “action to communicate” as required by RSA 173-B:1, IV. We disagree.
In essence, the defendant asks us to rewrite the statute. The defendant’s argument that “contact” requires that the defendant “be the actor not only [372]*372in the creation of the message, but in the conveyance of it to the protected person,” is fatally undermined by the legislature’s definition of the term “contact” in RSA 173-B:1, IV. The statute provides that “any action to communicate with another either directly or indirectly” constitutes contact. See RSA 173-B:1, IV (emphases added).
Additionally, the defendant incorporates a narrow “conveyance” requirement in his proffered interpretation of “contact,” that would require that the defendant deliver the message directly to “the protected person.” Although we agree with the defendant that “contact” requires more than merely creating a message, his limitations do not find support in the actual language chosen by the legislature, which requires only that a person act “either directly or indirectly” to “communicate with another.” See RSA 173-B:1, IV (emphases added). Further, in this case, the defendant did more than merely create a message. By posting messages addressing the victim on his public Facebook page, and directing the victim’s attention to his page, the defendant both created a message and took steps to convey it to the victim. To construe the statute as not encompassing the defendant’s conduct — writing a message addressing the victim and posting it in a public forum, but not personally conveying the message to the victim — would add limiting language that the legislature did not include. See Landry v. Landry, 154 N.H. 785, 788 (2007) (concluding that phrase “any property” requires broad interpretation of statute). Such a change to the statute’s language is not for this court to make. Id. “The legislature’s choice of language is deemed to be meaningful.” O’Brien v. N.H. Democratic Party, 166 N.H. 138, 143 (2014) (quotation and brackets omitted).
Moreover, it is significant that RSA 173-B:1, IV lists “any form of electronic communication” in its nonexhaustive list of “action[s] to communicate.” This reflects the legislature’s awareness that technological advances in communication — including e-mail and social media websites such as Facebook — provide a fertile environment for criminal behaviór and that “[s]ometimes, particularly in stalking and harassment cases, social media facilitates the crime.” Morrison, Passwords, Profiles, and the Privilege Against Self-Incrimination: Facebook and the Fifth Amendment, 65 Ark. L. Rev. 133, 136 (2012); cf. Beagle, Comment, Modem Stalking Laws: A Survey of State Anti-Stalking Statutes Considering Modem Mediums and Constitutional Challenges, 14 CHAP. L. Rev. 457,474 n.129 (2011) (observing that New Hampshire, like many states, added “electronic communications” to its general anti-stalking statute to address growing problem of cyberstalking).
Additionally, as we have previously recognized:
[373]*373the legislature intended RSA chapter 173-B to be construed liberally. See N.H.H.R. JOUR. 649 (1999). “It is the public policy of this state to prevent and deter domestic violence through equal enforcement of the criminal laws and the provision of judicial relief for domestic violence victims.” Id. at 648. A broad interpretation of the statute comports with the legislative purpose to “preserve and protect the safety of the family unit for all family or household members by entitling victims of domestic violence to immediate and effective police protection and judicial relief.” Id. at 649.
State v. Kidder, 150 N.H. 600, 603 (2004). Therefore, a broad interpretation of “any action to communicate” comports with the legislative purpose of RSA chapter 173-B: to provide those who seek protective orders pursuant to the statute with “immediate and effective police protection and judicial relief.” Id.
Our interpretation finds support in cases from other jurisdictions in which defendants have been held accountable for posting messages on the internet. See Baughman, Friend Request or Foe? Confirming the Misuse of Internet and Social Networking Sites by Domestic Violence Perpetrators, 19 WlDENER L.J. 933, 959-60 (2010) (discussing cases where courts have found liability for internet communications). For example, in O’Leary v. State, the District Court of Appeal for the First District of Florida upheld a trial court’s finding that a defendant “sent” a threatening statement to one of his relatives and her romantic partner by posting it on his own Facebook page. O’Leary v. State, 109 So. 3d 874, 877 (Fla. Dist. Ct. App. 2013). The defendant in O’Leary identified his victims by name, and his. Facebook page was “accessible by any member of the public who wanted to view [it].” Id. at 875. A Facebook friend of the defendant read the post and informed the victims about it. Id. at 874-75. Although the defendant in O’Leary claimed that “he ‘sent’ nothing because he neither asked anyone to view the posting on his personal Facebook page, nor addressed the posting to anyone,” the Court of Appeal rejected that argument, stating that “a common sense review of the facts suggests that [the defendant] has done more than he contends.” Id. at 877. The court ruled that “by the affirmative act of posting the threats on Facebook, even though it was on his own personal page, [the defendant] ‘sent’ the threatening statements to all of his Facebook friends ...” Id. The court explained:
When a person composes a statement of thought, and then displays the composition in such a way that someone else can see it, that person has completed the first step in [sending a message]. . . .
[374]*374[The defendant] reduced his thoughts to writing and placed this written composition onto his personal Facebook page. . . . Given the mission of Facebook, there is no logical reason to post comments other than to communicate them to other Facebook users. Had [the defendant] desired to put his thoughts into writing for his own personal contemplation, he could simply have recorded them in a private journal, diary, or any other medium that is not accessible by other people.
Id. We recognize that, unlike the victim in this case, the recipient of the threat in O’Leary was a Facebook friend of the defendant, and he received the post by way of his Facebook News Feed. However, we find that, given the circumstances in this case — that the defendant directed the victim to his Facebook page — the O’Leary court’s rationale applies with equal force here: there is “no logical reason” for the defendant to post statements directed to the victim on Facebook other than “to communicate them.” Id. Had the defendant desired merely “to put his thoughts into writing for his own personal contemplation,” and not wished them to be communicated to the victim, he could have written his thoughts in “any other medium that is not accessible by other people.” Id.
Similarly, in Rios v. Fergusan, a Connecticut court upheld a restraining order against a defendant after he “posted a video on YouTube in which [the defendant] brandished a firearm in a rap song in which he states that he wants to hurt the applicant, to shoot her and to ‘put her face on the dirt until she can’t breathe no more.’ ” Rios v. Fergusan, 978 A.2d 592, 595 (Conn. Super. Ct. 2008). The court observed that “[the defendant’s] YouTube video [was] more than the mere posting of a message on an open Internet forum . . . [because] he specifically targeted his message at [the victim] by threatening her life and safety.” Id. at 599-600 (emphasis added). Further, the court explained, the defendant “posted the video on an Internet medium that can be disseminated world-wide, but the content of the video establishes that he was purposefully directing it to [the victim] . . . .” Id. at 601. (emphases added).
We find these cases to be instructive. In this case, as in O’Leary, the defendant chose to make his page “public,” meaning his page was “available to anyone, even to people without an account on [Facebook].” Diss, supra at 1844 n.17. If, as he asserted to Officer Chadwick at the time of his arrest, the defendant was only “expressing his feelings,” he could have chosen to make his page private, or recorded his thoughts in “any other medium that is not accessible by other people.” O’Leary, 109 So. 3d at 877; see Ehling, 961 F. Supp. 2d at 662 (discussing Facebook privacy options).
[375]*375Further, in several posts that the defendant wrote after he received the restraining order, he stated that he was aware that the victim was reading his Facebook page:
[S]inee you would have to choose to look at the things I say to you on Facebook, that. . . means my butt is covered.
The document I have here does not mention my Facebook wall. You lose again . .. HA HA.... [Y]ou mentioned your knowledge of [my Facebook page] in your complaint, yet did not say not to talk about you on here.
Since I know you have been viewing my wall for quite some time, I win.
These posts demonstrate that the defendant, like the defendant in O’Leary, had “no logical reason to post comments other than to communicate them” to the victim. O’Leary, 109 So. 3d at 877.
In addition, we note that most of the defendant’s posts would have been meaningless to any reader other than the victim. For example, the defendant instructed the victim to take certain actions and say specific things “on Friday,” referring to the final restraining order hearing, and referenced specific details of the victim’s complaint. The content of these posts shows that the defendant, like the defendant in Rios, “specifically targeted his message at [the victim],” and “purposefully directed]” his posts to her. Rios, 978 A.2d at 600; see also Baughman, supra at 960-61 (discussing case in which court held that content of MySpace posts revealed that defendant intentionally communicated a public message on MySpace to specific victim).
Finally, the defendant himself acknowledged in his motion to dismiss that “it would violate the [stalking] statute if ... he had the intent [to make contact] and was in a place where he knew [the victim] might be.” Although during oral argument the defendant attempted to distinguish his Facebook posts from “standing out on the street corner where [he] might know [the victim] is going to be present and shouting out” to the victim — an act that he concedes would constitute an “action to communicate” and, thus, contact — we find the defendant’s posited scenario to be materially equivalent to the situation in this case. In both circumstances, the defendant’s contact with the victim is calculated, not fortuitous. The defendant’s posts reveal [376]*376that he was aware that the victim had been reading his posts on Facebook. We discern no meaningful difference between the defendant posting messages on Facebook with both the purpose and effect of communicating a message to her, and the defendant positioning himself on a street corner with the knowledge and expectation that the victim would pass by, and then shouting to her. For all of these reasons, we conclude that the defendant’s conduct was sufficient to constitute an “action to communicate” pursuant to RSA 173-B:1, IV.
We next address the defendant’s argument that, because the victim “voluntarily retrieved” rather than “merely received” the defendant’s messages when she searched for his Facebook page, he did not “contact” her in violation of the restraining order. We disagree.
First, nothing in the language of RSA 173-B:1, IV or RSA 633:3-a, 1(c) addresses the actions of a victim or the recipient of a message, nor states that they have a bearing on the issue of “contact.” ‘We will not consider what the legislature might have said or add language that the legislature did not see fit to include.” Dor, 165 N.H. at 200. Although we can envision circumstances in which a protected person’s conduct could impact the “contact” analysis — for example, if the protected person, without enticement, sought out the restrained person — this case does not present such a circumstance.
Moreover, to deny the victim in this, ease protection under the stalking statute would frustrate the statute’s purpose and thwart the intent of the legislature. The legislature passed RSA chapter 633:3-a with a focus upon protecting individuals from “domestic violence and problems of like gravity, such as threatening strangers and obsessive former lovers,” and in recognition of the fact that “[h]arassing and threatening behaviors toward innocent people is a serious problem.” Fisher v. Minichiello, 155 N.H. 188, 195 (2007) (Dalianis, J., concurring) (quotation omitted). Acknowledging that “[s]talking is a part of [domestic violence],” the legislature enacted RSA chapter 633:3-a in response to the “wide spread need in New Hampshire for legislation to allow the police to interfere before a domestic violence situation escalates into violence.” Id. (quotations omitted). We conclude, therefore, that interpreting RSA chapter 633:3-a to deny protection to a victim who has viewed publicly available Facebook posts and alerted the police to the threatening messages would frustrate the purpose of the stalking statute. Were we to conclude otherwise, the incongruous and potentially dangerous result would be — as the defendant himself observed in his Facebook posts directed to the victim — that the restraining order, rather than restraining the threatening behavior of the defendant, would make the victim “the one in cuffs.”
[377]*377Notably, the defendant does not cite, and we are unable to find, any case law supporting his assertion that the victim’s affirmative act of finding and reading his Facebook posts operates to bar his conduct from constituting “contact” pursuant to RSA 173-B:1, IV. In Commonwealth v. Butler, 661 N.E.2d 666, 667 (Mass. App. Ct. 1996), the court concluded that an indirect communication can constitute contact by the defendant, despite a victim’s affirmative act. Although the case did not involve online conduct, the court held that the defendant had violated a restraining order by anonymously sending flowers to the victim. Id. at 666-67. In that case, the restraining order provided that the defendant was “not to contact [the victim] either in person, by telephone, in writing, or otherwise, either directly or through someone else.” Id. at 666. The victim then received flowers at her home, with a card that gave the sender’s name as “requested withheld.” Id. Suspecting that the defendant was the sender, the victim called the florist and confirmed that the defendant had sent them. Id. The defendant had not given the florist his name, address, or telephone number, and wanted no name on the card. Id. Nonetheless, the court concluded that the defendant violated the “no contact” order, because he “achieved a communication with [the victim] amounting to ‘contact.’ ” Id. at 667. The court noted that “[the defendant’s] profession of anonymity merely invited inquiry” by the victim into the identity of the person who sent the flowers. Id.
Like the defendant in Butler, the defendant here “achieved a communication” with the victim indirectly. As in Butler, the defendant took deliberate steps to communicate with the victim while attempting to avoid culpability for violating the terms of the restraining order. As the victim testified, she would have had no “reason or desire to go look up [the defendant’s] Facebook page” if she had not received the defendant’s letter in which he told her that he was “speaking highly of [her] on [his] Facebook page,” or if her mother had not read the defendant’s posts and urged her to read them. Although the letter was sent prior to the issuance of the restraining order, and, therefore, its mailing did not, standing alone, violate the restraining order, it is critical in that it invited the victim’s later inquiry as to the nature of the defendant’s Facebook posts. See In the Matter of McArdle & McArdle, 162 N.H. 482, 487 (2011) (observing that acts committed by defendant prior to grant of restraining order are relevant to court’s inquiry as to whether recent acts pose a credible present threat to safety). The victim was “alarmed” by the defendant’s letter referencing his Facebook posts, and was urged by her mother to read the posts; it strains credulity to expect that the victim — or any person in her position — would refrain from ensuring her own safety by searching for and reading the defendant’s public Facebook page. Thus, just as the defendant in Butler [378]*378“invited inquiry” into who had anonymously sent the flowers, Butler, 661 N.E.2d at 667, the defendant here, by referring to Facebook posts in his letter, “invited inquiry” by the victim into what his posts said. The reasoning of Butler is equally applicable in the context of the internet. We conclude that the victim’s affirmative act of viewing the defendant’s Facebook page does not preclude the defendant’s conduct from constituting “contact” as used in RSA 173-B:1, TV.
The legislature expansively defined “contact” in RSA 173-B:1, IV to include “any action to communicate with another either directly or indirectly.” See RSA 173-B:1, IV (emphases added). Here, the defendant, having alerted the victim to his Facebook posts in his earlier correspondence, and believing that the victim was reading his posts, continued to post messages directed to her on his public profile page after he had been served with the restraining order. The victim’s mother urged her to read the posts due to “the extent and the severity and the vulgar use of words” about her daughter. In sum, examining the defendant’s conduct “in the context of all the evidence, not in isolation,” Germain, 165 N.H. at 355 (quotation omitted), and “considering all the evidence and all reasonable inferences therefrom in the light most favorable to the State,” id. at 354-55 (quotation omitted), we conclude that the defendant has not met his burden to demonstrate that no rational trier of fact could have found beyond a reasonable doubt that the defendant’s posts on his public Facebook page constitute “contact” in violation of the protective order. Accordingly, we conclude that the trial court properly denied the defendant’s motion to dismiss the stalking charge.
In so ruling, we need not decide whether a public Facebook post, standing alone, is sufficient to constitute “contact” pursuant to RSA 173-B:1, IV. Moreover, although we are mindful that Facebook posts, under some circumstances, may constitute protected speech under the First Amendment of the United States Constitution, see Elonis v. United States, No. 13-983 (U.S. argued Dec. 1, 2014) (addressing question of whether defendant’s conviction for posting threats on his personal Facebook page violates his free speech rights under the First Amendment), no First Amendment argument was advanced by the defendant in this case, nor does he contend that his Facebook posts served some legitimate purpose aside from communicating with the victim. Cf. RSA 633:3-a, 11(a).
B. Witness Tampering
We turn next to the defendant’s argument that the State presented insufficient evidence on the charge of witness tampering. See RSA 641:5. Pursuant to RSA 641:5, a person is guilty of witness tampering if, “[bjelieving that an official proceeding ... or investigation is pending or [379]*379about to be instituted, he attempts to induce or otherwise cause a person to ... [t]estify or inform falsely.” See RSA 641:1, II (2007) (defining “official proceeding”). There is sufficient evidence of witness tampering if the State proves that “the testimony the defendant sought to induce was in fact false, and that the defendant acted purposely to induce [the victim] to testify to something which the defendant believed was false.” State v. DiNapoli, 149 N.H. 514, 516-17 (2003).
The defendant first argues that the State failed to prove that the defendant acted purposely because it did not show that the defendant believed that the testimony he attempted to induce from the victim — telling the judge that she was “all set with [the defendant],” “[was] mistaken,” and no longer wanted the restraining order — would actually be false. The defendant asserts that an individual like himself — with an erroneous or delusional view of reality — might subjectively believe that the statements that he asked the victim to make to the judge or the police were actually true, and, therefore, the State could not establish that he acted purposely with knowledge that the testimony was false. The State counters that the defendant was aware that the victim feared him, and, therefore, that he knew that asking her to tell the judge that she no longer wanted the protective order would require her to testify falsely.
In State v. DiNapoli, we “assume[d] without deciding that the State had to prove that the testimony the defendant sought to induce was, in fact, false” to convict the defendant of witness tampering. Id. at 516. Making the same assumption here, we must, as we did in DiNapoli, “consider whether the State presented sufficient evidence to support a finding that the testimony the defendant attempted to induce was false and that he knew it was false.” Id.
“Because persons rarely explain to others the inner workings of their minds or mental processes, a culpable mental state must, in most cases, as here, be proven by circumstantial evidence.” Id. (quotation omitted). The jury is entitled to infer the requisite intent from the defendant’s conduct in light of all the circumstances in the case because “conduct illuminates intent.” Id. at 516-17 (quotation and brackets omitted). Thus, here, we look to the circumstances surrounding the victim’s interactions with the defendant to determine whether a reasonable jury could have found that he acted purposely, meaning that he believed the statements he sought to induce with his Facebook posts to the victim were, in fact, false.
The victim testified that she had no relationship with the defendant beyond her interactions with him at work, and that she was alarmed by his letters. The defendant was made aware of the victim’s complaints and concerns about his behavior when he received the stalking warning letter, [380]*380a no-trespass notice from the victim’s employer, and the restraining order against him. Further, the defendant himself acknowledged in his first letter that the victim was not interested in him, writing, “It’s not as if you are actually into me, or you wouldn’t be with someone else.” Given these facts, we find that a rational jury could have concluded that the statements that the defendant told the victim to make were in fact false and that the defendant believed that they were false. See id. (considering testimony and entirety of defendant’s conduct to find sufficient evidence of witness tampering); State v. Baird, 133 N.H. 637, 641 (1990) (considering “the totality of the evidence presented” when evaluating falsity in witness tampering charge).
The defendant next argues that the State did not prove that the defendant understood that he was asking the victim to give false testimony. The defendant contends that he merely asked the victim to “drop the charge,” which he asserts is very different from asking her to testify falsely. The State counters that the defendant mischaracterizes his Facebook posts and that, in fact, the defendant sought to intimidate the victim into testifying falsely at the upcoming hearing. We agree with the State.
In support of his argument, the defendant relies upon Rantala v. State, 216 P.3d 550, 556 (Alaska Ct. App. 2009), claiming that he, like the defendant in that case, merely asked the victim to “tell the authorities (whether she testified or not) that she did not wish to pursue” her complaint against the defendant. Rantala, 216 P.3d at 556. The defendant’s reliance on Rantala is misplaced. In Rantala, the defendant was charged with burglary after breaking into the home where he had previously lived with the witness. Id. at 555. Prior to the grand jury hearing on his burglary charge, the defendant told the witness that she did not have to testify if she had not been subpoenaed, and asked the witness not to pursue the case against him because he “believed that the burglary prosecution could not go forward” without her consent. Id. at 555-57. The court concluded that the defendant’s statements did not constitute a “request, or even a suggestion, that [the witness] lie about what happened” if the case went forward and she was subpoenaed to testify, and, therefore, he did not violate the witness tampering statute. Id. at 557.
In contrast, here, the defendant did not merely ask the victim not to pursue the case; rather, the defendant’s Facebook posts evidence that he was aware that the victim would be testifying at the final hearing on Friday, May 4, and that he sought to affect her testimony before the judge at that hearing. On the Saturday before the scheduled hearing, the defendant posted on Facebook repeatedly, .urging the victim to do and say specific things “on Friday.” For instance, the defendant wrote: “On Friday, you can [381]*381either tell the judge you are all set with me ... [o]r, you can drop all [the] charges”; “[Y]our options are to be all mine as of this Friday, or f**k off forever”; and “I want the order removed before Friday now.... You go tell the judge that you were mistaken.” Looking to “the inferential meaning of the [defendant’s] words and the context in which they were used,” Rantala, 216 P.3d at 557 (quotation and ellipsis omitted), a rational jury could infer that the defendant’s repeated reference to “Friday” indicated that he was asking the victim to make statements during the restraining order hearing scheduled for May 4. See Germain, 165 N.H. at 355 (explaining that jury may draw reasonable inferences from facts proved and facts found as a result of other inferences).
Additionally, the defendant’s Facebook posts reflect his understanding that the victim would be “under oath or affirmation” when she spoke at the restraining order hearing. See Black’s Law DICTIONARY 1704 (10th ed. 2014) (defining “testimony” as “[e]vidence that a competent witness under oath or affirmation gives at trial”). The defendant wrote that the victim would have to “lie under an oath of God,” and suggested that she could “go to jail for perjury.” Both statements show that the defendant understood that the victim would be testifying under oath at the restraining order hearing.
Finally, the defendant’s Facebook posts belie his contention that he merely asked the victim to drop the charges. He wrote, “On Friday, you can either tell the judge you are all set with me ... [o]r you can drop all [the] charges and become an honest woman.” (Emphases added). Because the defendant posited two distinct scenarios, and offered the victim a choice between testifying falsely and “drop[ping] all [the] charges,” a rational jury could have concluded that, “in light of all the circumstances,” the defendant understood that he urged the victim to give false testimony. See DiNapoli, 149 N.H. at 516-17. Accordingly, we conclude that the trial court did not err when it denied the defendant’s motion to dismiss the witness tampering charge.
Affirmed.
Dalianis, C.J., and Hicks, Conboy, and Lynn, JJ., concurred.