concur that the trial court I erred with respect to the first prong of the anti-SLAPP analysis. I further concur that Haight Ashbury Free Clinics, Inc. (HAFCI), failed to establish a probability of success on the merits of its third and fourth causes of action based on the protected activity alleged in paragraphs 31(o) and 31(p). I dissent, however, as to the majority opinion’s ruling that the third and fourth causes of action must be stricken in their entirety, as sought in appellants’ motion, even though only two of the 16 bases for liability had anything to do with the type of activity the anti-SLAPP (strategic lawsuit against public participation) statute was intended to protect. Such a result grossly miscasts the purpose of the anti-SLAPP law and grants a windfall to appellants.
I begin with a discussion of the second prong of the analysis for special motions to strike under the anti-SLAPP statute, because it is critical to understanding the problem that arises when a cause of action is based on allegations of both protected activity and unprotected activity.
*1556A. Determining Probability of Success on the Merits in a Mixed Cause of Action
Once a defendant has made a threshold showing that the plaintiff’s cause of action arises from the defendant’s free speech or petitioning activity, the burden shifts to the plaintiff to establish a probability of prevailing on the claim. (Equilon Enterprises v. Consumer Cause, Inc. (2002) 29 Cal.4th 53, 67 [124 Cal.Rptr.2d 507, 52 P.3d 685].) Where, as here, the cause of action is based on allegations of nonprotected activity as well as allegations of protected activity, the question arises whether the plaintiff might satisfy this burden by showing it could prevail based on any of the allegations underlying the causes of action, or whether it must show it could prevail based on the allegations of protected activity alone.
The majority opinion adopts the rule set forth in Mann v. Quality Old Time Service, Inc. (2004) 120 Cal.App.4th 90, 106 [15 Cal.Rptr.3d 215] (Mann).1 In Mann, the court ruled that, where a cause of action is based on both protected and nonprotected activity, the plaintiff needs to show in the second prong of the anti-SLAPP analysis only that it has a probability of prevailing on any part of its claim, whether it be by showing some merit to the allegations of protected activity or some merit to the allegations of nonprotected activity. (Mann, supra, 120 Cal.App.4th at pp. 100, 106.)2
*1557I cannot agree with Mann on this point. The upshot of Mann is that a plaintiff may freely make utterly meritless attacks on a defendant’s exercise of free speech and petition, if the plaintiff has also alleged some nonprotected activity that supports the cause of action. This is plainly inconsistent with the anti-SLAPP statute’s purpose of curbing lawsuits that chill the valid exercise of free speech. As the court in Peregrine Funding remarked in another context: “Where, as here, a cause of action alleges the plaintiff was damaged by specific acts of the defendant that constitute protected activity under the statute, it defeats the letter and spirit of section 425.16 to hold it inapplicable because the liability element of the plaintiff’s claim may be proven without reference to the protected activity.” (Peregrine Funding, Inc. v. Sheppard Mullin Richter & Hampton LLC (2005) 133 Cal.App.4th 658, 674 [35 Cal.Rptr.3d 31].)
Given the overriding statutory purpose, the language of the anti-SLAPP statute confirms that a plaintiff can save his claim of liability based on protected activity only if he establishes a modicum of merit to that particular claim. Subdivision (b)(1) of Code of Civil Procedure section 425.16 provides: “A cause of action against a person arising from any act of that person in furtherance of the person’s right of petition or free speech under the United States Constitution or the California Constitution in connection with a public issue shall be subject to a special motion to strike, unless the court determines that the plaintiff has established that there is a probability that the plaintiff will prevail on the claim.” (Italics added.) Thus, while stating that a “cause of action” targeting protected activity may be stricken, the statute allows the plaintiff to avoid that consequence by showing a probability of prevailing on “the claim.” (§ 425.16, subd. (b)(1), italics added.) Assuming, as we must, that the Legislature had something in mind when it used the word “claim” as opposed to “cause of action,” the reasonable reading of the statute is that the plaintiff must show a probability of prevailing on the “claim” of liability that is premised on the protected activity, not the “cause of action” on some other basis.
Furthermore, Mann itself leads me to believe that the view stated in that case is incorrect. The court in Mann concluded a plaintiff establishes that its cause of action has merit if it shows a probability of prevailing on “any part of its claim.” (Mann, supra, 120 Cal.App.4th at p. 106.) As mentioned, however, the anti-SLAPP statute requires that the plaintiff show a probability of prevailing not on “any part of its claim” (120 Cal.App.4th at p. 106) or any part of its cause of action, but particularly on “the claim” (Code Civ. Proc., § 425.16, subd. (b)(1)), referring to the allegations targeting protected *1558activity. In other words, the Legislature intended that the plaintiff demonstrate some minimal indication of merit to the allegations that target the particular activity the anti-SLAPP statute intended to protect.
The court in Mann also expressed its concern that a trial court not have to “engage in the time-consuming task of determining whether the plaintiff can substantiate all theories presented within a single cause of action and need not parse the cause of action so as to leave only those portions it has determined have merit.” (Mann, supra, 120 Cal.App.4th at p. 106.) This concern, however, is misplaced. If the plaintiff is required to demonstrate merit to his cause of action based on the allegations of protected activity, the trial court would not have to determine the merits of “all theories,” but only those premised on protected activity. Indeed, this may often be a simpler task than having to consider the merit of theories based on nonprotected activity. In the matter before us, for example, Mann would result in the trial court considering the merits of as many as 14 breaches of fiduciary duty based on nonprotected activity, rather than no more than two breaches of fiduciary duty based on protected activity.
Lastly, the court in Mann pointed out that “a defendant has other options to eliminate theories within a cause of action that lack merit or cannot be proven,” such as a motion to strike under Code of Civil Procedure section 436 or a motion for summary adjudication. (Mann, supra, 120 Cal.App.4th at p. 106.) The fact that a defendant may have other ways of removing the allegations of protected activity from the case, however, is immaterial. After all, it was the insufficiency of these other procedural mechanisms that motivated our Legislature to enact the anti-SLAPP statute in the first place. Forcing a defendant to go to the expense, time and effort to remove the allegations in other ways does nothing to curb the chilling effect on the exercise of free speech and petition rights, and is precisely what the antiSLAPP statute was intended to avoid.
Based on the language and purpose of the anti-SLAPP statute, a plaintiff should have to demonstrate, with admissible evidence, a probability that it would prevail on its cause of action arising from protected activity, based on its allegations of protected activity. (Indeed, HAFCI has not contended otherwise in this case.)
B. Effect of a Minority of Severable Allegations of Protected Activity
If a plaintiff establishes a probability of prevailing on its cause of action, based on its allegations of protected activity, the motion to strike will be denied. If, on the other hand, the plaintiff does not establish a probability of prevailing, the question becomes whether, as appellants contend, the entirety of the cause of action must be stricken without leave to amend.
*1559More specifically, in the context of a mixed cause of action, the following question arises: must a cause of action, based almost wholly on nonprotected activity, be stricken in its entirety, without leave to amend, where it was also based on much fewer and severable allegations of protected activity? Or, to put it slightly differently, may a defendant use the anti-SLAPP statute as a weapon to knock out permanently an entire cause of action that the plaintiff had supported with numerous allegations of nonprotected wrongdoing, merely because the plaintiff had also alleged (and, in this case, offered to delete) allegations of protected activity?
The answer must be no. Even if a plaintiff cannot show that it could prevail on its cause of action based on its allegations of protected activity, it makes neither good sense nor good policy to strike the cause of action in its entirety, with prejudice, where there may well be ample evidence of a breach of fiduciary duty based on nonprotected activity.3 After all, if the plaintiff had alleged one cause of action for breach of fiduciary duty based solely on nonprotected activity, and a second cause of action for breach of fiduciary duty based solely on protected activity, obviously only the second of those causes of action would be stricken. There is no reason to foreclose the equivalent result where the allegations have been combined in a single cause of action. Never was it the purpose of the anti-SLAPP statute to allow a defendant to force a trial court to throw out allegations that are neither subject to the statute nor posing any risk of chilling the exercise of constitutional rights of free speech and petition.
Where, as here, protected activity is alleged as an independent and alternative basis for liability in a cause of action, the inappropriate forfeiture of the entire cause of action might conceivably be avoided in one of two ways: (1) striking the entire cause of action, but permitting the plaintiff to amend the complaint solely to reallege the cause of action without the allegations of unsupported protected activity; or (2) striking just the allegations of protected activity for which the plaintiff has not shown a prima facie case. Both approaches accomplish the same equitable result.
1. Amendment Solely to Repeat Cause of Action Without Protected Activity
Permitting the plaintiff to amend the complaint solely to reassert the same cause of action, without the allegations that had targeted protected activity, accomplishes the purposes of the anti-SLAPP statute. The offending allegations are removed, and the case proceeds on the causes of action based on *1560activities not subject to the anti-SLAPP statute. To the extent the defendant has incurred attorney fees and costs on its motion to strike, they are recoverable under Code of Civil Procedure section 425.16, subdivision (c).
I am mindful that, in a vastly different context, it has been held that a plaintiff should not be allowed to amend its complaint after an anti-SLAPP motion has been granted. (Simmons v. Allstate Ins. Co. (2001) 92 Cal.App.4th 1068, 1073 [112 Cal.Rptr.2d 397] (Simmons); see also Sylmar Air Conditioning v. Pueblo Contracting Services, Inc. (2004) 122 Cal.App.4th 1049, 1054-1056 [18 Cal.Rptr.3d 882] [filing of first amended complaint pursuant to Code Civ. Proc., § 472 before the hearing on the anti-SLAPP motion to strike did not render the motion to strike moot].) Simmons is plainly inapposite.
In Simmons, the appellant had filed a cross-complaint alleging that the respondent conspired to force him out of business by filing frivolous lawsuits, waging a media war, and making defamatory statements and refusing to pay claims. (Simmons, supra, 92 Cal.App.4th at p. 1071.) In the published portion of the opinion, the court addressed the appellant’s request for leave to amend the cross-complaint. The court opined that “[allowing a SLAPP plaintiff leave to amend the complaint once the court finds the prima facie showing has been met would completely undermine the statute by providing the pleader a ready escape from section 425.16’s quick dismissal remedy.” (Simmons, at p. 1073.) “Instead of having to show a probability of success on the merits, the SLAPP plaintiff would be able to go back to the drawing board with a second opportunity to disguise the vexatious nature of the suit through more artful pleading. This would trigger a second round of pleadings, a fresh motion to strike, and inevitably another request for leave to amend.” (Ibid.) “By the time the moving party would be able to dig out of this procedural quagmire, the SLAPP plaintiff will have succeeded in his goal of delay and distraction and running up the costs of his opponent. . . . This would totally frustrate the Legislature’s objective of providing a quick and inexpensive method of unmasking and dismissing such suits.” (Id. at p. 1074, citation omitted.) The court concluded: “the omission of any provision in section 425.16 for leave to amend a SLAPP complaint was not the product of inadvertence or oversight.” (92 Cal.App.4th at p. 1074.)
The concerns articulated in Simmons would certainly not be implicated by allowing the plaintiff to reallege a cause of action, this time without the allegations of protected activity. Here plaintiff has already alleged nonprotected activity, as well as protected activity, as a basis for the cause of action. Permitting a plaintiff to reallege his cause of action based solely on the nonprotected activity he had already alleged would not give the plaintiff any new opportunity to disguise a vexatious lawsuit or compel another antiSLAPP motion; it would merely make the “punishment” fit the plaintiff’s *1561“crime.” While the Legislature might well have wanted to avoid granting leave to amend in the context and manner described by Simmons, there is no compelling reason to conclude it would frown on the reinstatement of previous allegations that had not been challenged by the defendant or, in any event, pose no threat to the purposes of the anti-SLAPP statute.
Although not on point factually, Nguyen-Lam v. Cao (2009) 171 Cal.App.4th 858, 870-873 [90 Cal.Rptr.3d 205] (Nguyen-Lam), is instructive. In Nguyen-Lam, the plaintiff sued the defendant for slander, and the defendant filed an anti-SLAPP motion. {Id. at p. 864.) As to the first prong of the anti-SLAPP analysis, the trial court concluded that the allegedly slanderous statement was protected under the anti-SLAPP statute, but as to the second prong of the analysis ruled that the evidence showed a probability the plaintiff would prevail on the claim that the defendant had slandered the plaintiff with actual malice. (Id. at pp. 865-866.) The problem was, the plaintiff had apparently not alleged actual malice in the complaint. {(d. at pp. 862, 865-866.) The trial court allowed the plaintiff to amend the complaint to allege actual malice and (effectively) denied the anti-SLAPP motion. {Id. at pp. 862, 866.)
The appellate court affirmed, distinguishing Sylmar and Simmons. Distinguishing Sylmar, the court noted that the plaintiff in Nguyen-Lam was not avoiding or frustrating a hearing on an anti-SLAPP motion. (Nguyen-Lam, supra, 171 Cal.App.4th at pp. 871-872.) The same must be said in this case regarding leave to reallege the cause of action based solely on nonprotected activity.
Distinguishing Simmons, the court in Nguyen-Lam explained that the amendment in Nguyen-Lam was merely to permit the plaintiff to allege actual malice for which there was evidence. Because the plaintiff demonstrated a probability of prevailing based on the evidence, the action was not a SLAPP, and therefore amendment of the complaint should not be precluded. (Nguyen-Lam, supra, 171 Cal.App.4th at p. 871.) Of most interest to the matter at hand, the court in Nguyen-Lam also rejected the notion that Simmons established a per se bar to amendments after rulings on anti-SLAPP motions, and suggested the legislative purpose of the anti-SLAPP statute should determine the extent to which amendment should be permitted. {Id. at pp. 872-873.) As stated ante, the availability of leave to reinstate a cause of action, without the allegations of protected activity, is consistent with the purposes of the anti-SLAPP statute.4
*15622. Striking Unsupported Allegations of Protected Activity
Perhaps the most direct and efficient solution is simply to strike the specific allegations of protected activity for which a plaintiff is unable to show a probability of success on the merits. In this way, without the need for any amendment to the pleading, the intent of the anti-SLAPP statute will be met, the plaintiff will be able to proceed on a cause of action not subject to the anti-SLAPP statute, and the defendant will not obtain a windfall.
In considering a different issue, the court in Mann remarked that “the anti-SLAPP procedure may not be used like a motion to strike under [Code of Civil Procedure] section 436, eliminating those parts of a cause of action that a plaintiff cannot substantiate.” (Mann, supra, 120 Cal.App.4th at p. 106.) As explained ante, however, the concern in Mann was that a trial court should not have to be saddled with the “time-consuming task of determining whether the plaintiff can substantiate all theories presented within a single cause of action.” (Ibid., italics added.) That concern does not arise where, as I propose, the trial court merely decides whether the plaintiff showed a prima facie case of liability based on the protected activity and, if not, strikes the allegations of protected activity.
My conclusion is also consistent with the language of the anti-SLAPP statute. The statute provides that a cause of action arising from protected activity is “subject to a special motion to strike” unless the plaintiff has shown a probability of prevailing on his claim. (Code Civ. Proc., § 425.16, subd. (b)(1).) Nowhere does it require the entirety of a cause of action to be stricken. Nor does it preclude striking just part of a cause of action or specific allegations. Moreover, the Legislature’s reference to a “cause of action” cannot mean that a court would be helpless to do anything other than strike the entirety of the cause of action, if it led to a result contrary to the anti-SLAPP statute’s purpose.
Indeed, my conclusion is fully supported by the purpose of the anti-SLAPP statute: to strike meritless assertions of liability based on protected activity, but only meritless assertions of liability based on protected activity, in order to curb the discouragement of the valid exercise of constitutional rights to free speech and petition.
I therefore conclude: where a single cause of action asserts liability based on a list of wrongful acts, some of which constitute protected activity under *1563Code of Civil Procedure section 425.16, subdivision (e), and some of which do not, and the plaintiff fails to demonstrate a probability of success based on the allegations of protected activity, the trial court should strike the unsupported allegations of protected activity from the cause of action and leave the balance of the cause of action intact.5
A petition for a rehearing was denied June 23, 2010. Needham, J., was of the opinion that the petition should be granted. Respondent’s petition for review by the Supreme Court was denied September 1, 2010, SI84232. Kennard, J., and Corrigan, J., were of the opinion that the petition should be granted.
The majority opinion does not rebut my analysis of Mann, but suggests I should not write about it, claiming that the “validity of the Mann analysis (or any other analysis) is not really before us” in light of HAFCI’s failure to try to satisfy its burden under the second prong of the anti-SLAPP analysis. (Maj. opn., ante, at p. 1554.) However, the majority quotes the rule set forth in Mann and embraces it as “[t]he test that ordinarily would be used.” (Maj. opn., ante, at p. 1554.) I believe it should not be. Moreover, my discussion of Mann is not for the purpose of deciding whether HAFCI met its burden on the second prong, but to help demonstrate the unreasonableness of striking the entirety of the causes of action—which the majority opinion directs the trial court to do. The majority opinion also notes that the Legislature has had the opportunity to amend the anti-SLAPP statute if it thought courts were interpreting it incorrectly. (Citing Olmstead v. Arthur J. Gallagher & Co. (2004) 32 Cal.4th 804, 816 [11 Cal.Rptr.3d 298, 86 P.3d 354] [where legislative history expressly acknowledged court decisions deciding an issue].) However, no published decision has ruled whether an entire cause of action must be stricken under the circumstances now before us. The issue has therefore not been brought to the Legislature’s attention—until now—and its silence up to this point tells us nothing. The matter is squarely before us in this case, the parties addressed the remedy at oral argument after our letter request, and we should deal with it.
The court in Mann stated: “Where a cause of action refers to both protected and unprotected activity and a plaintiff can show a probability of prevailing on any part of its claim, the cause of action is not meritless and will not be subject to the anti-SLAPP procedure. [H Stated differently, the anti-SLAPP procedure may not he used like a motion to strike under section 436, eliminating those parts of a cause of action that a plaintiff cannot substantiate. Rather, once a plaintiff shows a probability of prevailing on any part of its claim, the plaintiff has established that its cause of action has some merit and the entire cause of action stands. Thus, a court need not engage in the time-consuming task of determining whether the plaintiff *1557can substantiate all theories presented within a single cause of action and need not parse the cause of action so as to leave only those portions it has determined have merit.” (Mann, supra, 120 Cal.App.4th at p. 106, original italics.)
There is no provision in the anti-SLAPP statute requiring a plaintiff to demonstrate the merit of its allegations of nonprotected activity.
In Salma v. Capon (2008) 161 Cal.App.4th 1275 [74 Cal.Rptr.3d 873], Salma amended his cross-complaint before the trial court ruled on Capon’s anti-SLAPP motion to strike. After concluding that Salma’s original claims were subject to dismissal, we ruled that the amended claims would also be subject to automatic dismissal, because requiring the trial court to *1562analyze the amended claims would cause “the evils identified in Simmons” and run counter to the statutory purpose of early disposition of claims. (Id. at pp. 1294-1295.) We did not consider the propriety of permitting a limited amendment under the circumstances of the matter now before us, which does not present the evils with which we were concerned in Salma.
Tellingly, the striking of paragraphs 31(o) and 31(p) is what appellants requested before filing their motion to strike, and what respondent agreed to shortly after it was filed.