Fetishistic Textualism

An Example from Civil Procedure

The New Digest is delighted to announce that Jeremy M. Christiansen is joining our masthead. Professor Christiansen will start at the Regent University School of Law in Virginia Beach, Virginia as an Associate Professor this summer.

We should give textualism its due for reminding the American judicial system that text matters. But I have grown concerned over the years that textualism is becoming a meme of itself. And it is obscuring the judicial function, much like Mike Campbell’s slide into bankruptcy: “gradually, then suddenly.” Examples are increasing in number and frequency. The Supreme Court claims that in reviewing statutes, the Justices do not sit “as a panel of grammarians.” Yet in the Year of Our Lord 2025, you’d be forgiven for thinking that the Justices and lower courts do precisely that—with oral arguments, briefs, and opinions all too often fixated on minutiae of the role of past participles, last antecedents, and present perfect tense and their outsized impact on the meaning of federal statutes. Other examples are not hard to come by, but I’d like to focus on a phenomenon I’ve encountered with alarming frequency; a fetishistic pride in insisting that the “plain” or “unambiguous” meaning of a legal text requires some eye-brow-raising result, divorced from any plausible account of legal rationality. I will take an example from civil procedure, where in my view, a more classically oriented approach to thinking about the law could provide a more rational and common-sense interpretation.

Does Federal Rule of Civil Procedure 41(a)(1)(A)(ii) permit a plaintiff and a defendant to agree to dismiss particular claims within a lawsuit without a court order? Here is the relevant text (including 41(c), which will become relevant in the discussion):

(a) Voluntary Dismissal.

(1) By the Plaintiff.

(A) Without a Court Order. Subject to Rules 23(e), 23.1(c), 23.2 , and 66 and any applicable federal statute, the plaintiff may dismiss an action without a court order by filing:

(i) a notice of dismissal before the opposing party serves either an answer or a motion for summary judgment; or

(ii) a stipulation of dismissal signed by all parties who have appeared.

* * *

(c) Dismissing a Counterclaim, Crossclaim, or Third-Party Claim. This rule applies to a dismissal of any counterclaim, crossclaim, or third-party claim. A claimant's voluntary dismissal under Rule 41(a)(1)(A)(i) must be made:

(1) before a responsive pleading is served; or

(2) if there is no responsive pleading, before evidence is introduced at a hearing or trial.

If you research the question, you will find a unanimous opinion among federal appellate courts—the Rule limits a plaintiff’s ability to dismiss via stipulation without a court order, to the plaintiff’s entire case, and not individual claims. Importantly for our purposes here, respected treatises such as Moore’s Federal Practice have observed that this holding “is usually justified by resort to textual interpretation,” as have a number of the cases reaching this conclusion. If you survey the cases dealing with this issue, they are insistent that the outcome is commanded by the “plain text” of the Rule. I want to examine one of these text-driven decisions, Esteva v. UBS Financial Services Inc., a 2023 opinion from the U.S. Court of Appeals for the Eleventh Circuit, to unpack just how strange the unanimous reasoning of the courts has become here. Esteva arose in the bankruptcy context, so I have to beg the reader’s indulgence as I provide some background on bankruptcy cases and related jurisdictional issues to set the stage.

A bankruptcy case is concerned with the sorting out of a debtor’s debts viz. financial claims various creditors may have against the debtor. A bankruptcy can naturally, and usually does, involve many parties. Within the bankruptcy case, parties often file subsidiary lawsuits called “adversary proceedings” which otherwise look like a normal civil action, but they are housed underneath the umbrella of the broader bankruptcy case.

As those familiar with federal jurisdiction will know, there is a baseline statutory rule in federal law that dictates when a lawsuit is sufficiently “final” for a party to file an appeal as of right under the standard jurisdictional statute for appeals, 28 U.S.C. § 1291. This baseline test is known as the “final judgment rule.” For a federal appellate court to have jurisdiction over an appeal in the ordinary course, there must be a final judgment. This rule is usually articulated in common-sense language—a judgment is final when there is nothing left for the district court to do but enter judgment. Thus, an order that is interlocutory, such as the denial of summary judgment or an order disposing of less than all claims as to all defendants, is not “final” and an attempt to appeal it as a matter of right (rather than seeking alternative discretionary grounds for appeal) will be met with a stiff jurisdictional barrier at the appellate level. Bankruptcy cases, because they frequently involve so many parties and claims, can be a special kind of hell for this very reason—a litigant may have to wait a very long time to appeal, and, due to a doctrine that I will not get into here known as “equitable mootness,” it can be disastrous to wait to appeal to the end of a bankruptcy proceeding.

In Esteva the debtor commenced an adversary proceeding against a defendant bank to recover frozen funds the bank held in an account. The adversary complaint alleged four claims, and the bank countersued with four counterclaims of its own. The debtor ultimately moved for summary judgment on three of the four counts in its own adversary complaint, partial summary judgment on the fourth, and summary judgment on all four of the defendant’s counterclaims. The bankruptcy court granted the motion, but because the debtor sought only partial summary judgment on its fourth claim, the claim as a whole remained in the case as the sole issue for trial.

The defendant went ahead and appealed to the federal district court (as one can often do in bankruptcy settings), which affirmed, and then the defendant filed for an appeal before the Eleventh Circuit. Because of the still-pending claim unresolved on summary judgment, the Eleventh Circuit (rightly) became concerned it may lack jurisdiction and so directed the parties to brief the issue. Just before oral argument in that appeal, the parties filed a joint motion to supplement the record to include a joint stipulation they filed pursuant to a Bankruptcy Rule that mirrors Federal Rule of Civil Procedure 41(a)(1)(A)(ii), in which they agreed to dismiss the final remaining claim.

The Eleventh Circuit unanimously dismissed the appeal for lack of appellate jurisdiction. It held that “[t]hough the parties attempted to finally resolve the adversary proceeding by stipulating to a voluntary dismissal of the only remaining claim, their stipulation did not achieve its purported end because it was invalid upon filing.” Why? The court reasoned that Rule 41(a)(1)(A)(ii) “unambiguously” refers to an “action” not particular “claims” and thus does not authorize anything less than the dismissal of the whole suit by the plaintiff. The panel noted that the Rule drafters knew how to distinguish between “actions” and “claims,” using the term “claims” in Rule 41(b) and (d), for instance. So “action” must mean “whole lawsuit.” Therefore, the unresolved claim was still pending, rendering the judgment non-final and preventing the appellate court from obtaining jurisdiction and reaching the merits of the parties’ dispute. The panel tried to reassure the parties that all was not lost because they could seek to certify an otherwise non-final judgment through Federal Rule of Civil Procedure 54(b) or seek to amend the claim out of the pleadings via Rule 15. But they had to go back and try again.

One can sense the panel’s own need to defend the result reached. The panel reached for high principles to assure their readers that this was no small matter, but apparently one that involves some of the deepest questions of our constitutional structure: “[t]he result we reach today is not some technical application of a silly or obscure rule. Congress has the constitutional authority to define the jurisdiction of the lower federal courts, and, once the lines are drawn, limits upon federal jurisdiction … must be neither disregarded nor evaded.” In other words, because the parties’ attempted joint stipulation would also bear on the jurisdiction of the appellate court, this was a very serious matter.

Methinks the court doth protest too much. With all respect to our federal jurists, it seems to me that the rise of textualism has some share of the blame for a silly disposition here. Below, I offer a modest rebuttal perspective that is respectful of the text of the rule, but more holistic and concordant with the classical approach to law. Specifically, the presuppositions of textualists blind them, short-circuiting the legal reasoning process, quick to find a word, a comma, a definition, that plainly resolves the litigation. A fetish for the lex of the law, without an appreciation for the ius of the law. Classically, law consists not only in the lex or written text of a statute, rule, or regulation, but also ius—background, often unwritten, law and legal principles, purposes, telos, of rationality, justice and so forth that inform the lex and show its true meaning in particular circumstances.

Let’s consider just what Rule 41(a)(1) is aimed at doing: setting up a mechanism whereby, without the need to ask the district court’s permission, the plaintiff can secure the defendant’s agreement to dismiss “an action,” whatever an “action” is. Why is it singling out an option to dismiss an “action” without a court order? What is the significance of that? The Rule also has a provision for dismissal with a court order. If you have not litigated much, you might not appreciate just how useful this mechanism could be. Having to move the court to dismiss something, even if both sides agree to it, is costly and risky. It is costly because, unlike a mere notice of stipulated dismissal, which operates on its own terms, dismissal via court order requires you to brief and persuade the court, and it might be opposed by the defendant, meaning more briefing. Even assuming very modest billable rates of the relevant attorneys’, say $500 an hour, the cost of such a motion could run into a moderate five-figure endeavor rather easily, to say nothing of a case involving large clients and large law firms where the rates will be twice, three times, or even four times that amount per hour. Moreover, the elimination of even a single claim can drastically reduce the total cost of trial—trials, dear reader, are so unbelievably expensive I will not even give you a figure because you might not believe me. The additional downside of moving the court is that there is risk the court will say no. The basis of dismissal with a court order under Rule 41 is, “on terms that the court considers proper.” This is a purely discretionary standard, and believe it or not, sometimes district court judges have their own view of the litigation and how it ought to proceed, the parties be damned. So you might spend the money on the motion for nothing. This is frustrating as a lawyer, and very frustrating for clients whose rights are actually being adjudicated. This context can help the reader understand why the Eleventh Circuit telling the parties that Rules 54(b) and 15 might offer alternative paths to appeal a non-final judgment and/or remove a claim from the case ring hollow—both are discretionary with the district court, require meeting particular standards for being granted, and a denial of either would not itself be immediately appealable (or in the case of Rule 54, appealable at all). Rule 41 offers a uniquely efficient vehicle for dismissing an “action” because you are not asking permission or persuading the court. You get to just inform the court that the “action” is dismissed, and it is so.

Now, let’s think about the interpretive framework of the Rule. The Federal Rules of Civil Procedure helpfully identify how they are supposed to be interpreted and applied, that is, the ius and the telos of the Rules has been spelled out for us. Rule 1 states: “[These Rules] should be construed, administered, and employed by the court and the parties to secure the just, speedy, and inexpensive determination of every action and proceeding.” Here I can digress a bit to discuss a favorite retort of the textualist—purpose matters, but purpose is revealed in a legal text only insofar as the text advances the purpose and no further, because not every purpose is to be advanced at all costs. This is a straw man. No one says that the purposes of justice, speed, and inexpensiveness are to be advanced at all costs; just that the text should be read, and can only be read, in harmony with the purpose. The classical view is that law is not exhausted by the text, and that even the meaning of the text cannot be understood without an appreciation of what it is supposed to be accomplishing. Approaching the question of interpreting the procedural rules of civil litigation in federal court, we should always ask ourselves, does the interpretation reached advance or detract from the just, speedy, and inexpensive resolution of civil litigation? That no one asked this question, at least explicitly, in Esteva should give us pause.

Let’s give some credit where it is due to the Eleventh Circuit by acknowledging that the term “action” generally refers to a whole suit. Generally, but not always. I suspect that the belief that text is exhaustive of law can lead a jurist to short-circuit and fail to fully engage, because there might be a quick textual answer that seems right at first blush, but gets no further scrutiny. Is it really the case that “action” always refers to an entire suit and not individual claims? Turns out, the answer is no. Let’s consider standard statutes of limitations, where we routinely find a reference to an “action” or “actions.” It appears fairly common, in fact, for the term “action” to refer to individual claims, rather than an entire lawsuit in this context. The Federal Tort Claims Act’s statute of limitations provision provides, “every civil action commenced against the United States shall be barred unless the complaint is filed within six years after the right of action first accrues.” Similarly, the District of Columbia’s general statute of limitations provides that “actions for the following purposes may not be brought after the expiration of the period specified below from the time the right to maintain the action accrues,” then lists off a dozen types of actions with relevant time limitations, “for the recovery of damages for an injury to real or personal property—3 years”; and so on. If we apply the theory that “action” means the whole lawsuit to these limitations statutes, then one untimely claim means the whole “action” is barred. But every lawyer and every judge knows that is wrong; that is not how a limitations statute works. Despite saying that an “action” is barred if a right of action accrued outside the limitations period, it means that the cause of action is barred. And that makes sense because, in point of fact, these statutes use the word “action” to mean “cause of action.” This usage is common enough, even if less precise than it could be. But the take-home is that the analysis of Rule 41 is not so simple; we cannot just observe that “it says action not claim.”

Furthermore, as Justinian reminded us centuries ago, “[i]t is not lawyer-like practice to give judgment or to state an opinion on the basis of one particular part of a statute without regard to the whole.” (Digest, Book I, 3.24). The Eleventh Circuit, while considering some parts of the Rule, such as subsections (b) and (d) which appear to distinguish between “actions” and “claims” curiously omitted any analysis of Rule 41(c), which states, “[t]his rule applies to the dismissal of any counterclaim, crossclaim, or third-party claim.” By the panel’s own reasoning, if an “action” refers to a whole suit and a “claim” as in Rule 41(b) and (d) refers to individual claims, it follows that Rule 41(c)’s references to “any” counterclaims, crossclaims, and third-party claims refers not to the entirety of the countersuit, but to individuals counterclaims etc. within that party’s countersuit. In other words, it does appear that Rule 41(c) permits a defendant-counterclaiming to do exactly what plaintiffs are supposedly not allowed to do. This presents us with a problem, indeed, an absurdity. This situation is not unlike Green v. Bock Laundry Machine Co.. The literal reading of the text in that case would have mandated admission of impeaching convictions against plaintiffs in civil cases without any consideration of prejudice, while requiring a prejudice balancing before the same evidence could be introduced against the defendant. This forced all the Justices to opt for what they saw as a sub-optimal interpretation of the text to avoid a blatant absurdity, even though they disagreed on how to resolve it. There is a lesson for us. Even if the term “action” barely sustains the narrower meaning of “cause of action,” it is reasonable to ascribe it such a meaning to avoid gross absurdity; by parallel, Justice Scalia’s concurrence in Green would have narrowed the term “defendant” to “criminal defendant” to avoid a similar absurdity.

The Eleventh Circuit’s approach has even more absurdity lurking within it, however. The Eleventh Circuit’s opinion frankly acknowledges an exception in its case law, one that does not easily square with the court’s insistence that “action” means “entire lawsuit.” The court conceded that it has long interpreted Rule 41(a) to permit voluntary stipulated dismissals of all claims against only a particular defendant in multi-defendant proceedings. But that is obviously not the “action” in the sense the Eleventh Circuit ascribes to that term in Esteva. Dismissing some defendants entirely but not others necessarily eliminates only claims from the case, not the “action,” if we are tracking the Eleventh Circuit’s rationale. The outcome of Esteva measured against this exception only highlights that the Eleventh Circuit has rendered the statute more incoherent and inconsistent. Consider this hypothetical. Plaintiff A sues Defendants A1, A2, and A3 for breach of contract, breach of fiduciary duty, and unjust enrichment, respectively. Plaintiff B sues Defendant B with the same causes of action. Under the Eleventh Circuit’s approach, Plaintiff A can, without a court order, reach an agreement to dismiss its claims against Defendants A2 and A3, such that the case that goes to trial is only about Plaintiff A versus Defendant A1 on a breach of contract claim. Plaintiff B cannot do the same. Plaintiff B has to move the court to achieve this result, and spend the money to do so. Plaintiff B’s only option under Rule 41 is to dismiss his entire lawsuit against Defendant B. None of this makes any sense.

At minimum, Rule 41 is ambiguous (although I’m not even sure it is that). And “when there is an ambiguity in a statute, that sense is to be preferred which avoids an absurdity, especially when by this method the intendment of the act is also secured.” (Digest, Book I, 3.19). As we’ve noted, the “intendment” of the Rules of Civil Procedure is to secure the just, speedy, and inexpensive resolution of civil matters, purposes served by an interpretation that reads the Rule as permitting both plaintiffs and defendants (just) to dismiss particular claims from the case without a court order (speedy and inexpensive) by all agreeing to do so (agreement being both just, speedy, and inexpensive).

Equally puzzling in Esteva is the lack of any affirmative rationale for why the rule drafters might have chosen this peculiar scheme for voluntary dismissals without a court order. The panel implicitly concedes that there is nothing per se wrong or irrational or problematic about the parties achieving this same outcome via other means like Rule 15, seeking district court approval. So the panel does not identify anything wrong with the interpretation posited by the parties in the case that would permit them to dismiss the claim. At the same time, however, the panel offers no explanation for why the rule they are endorsing makes sense. Law is an ordinance of reason, so this is perplexing indeed. And that brings us to the Eleventh Circuit’s paean to the limited jurisdiction of the federal courts. I find this component of the opinion among the least persuasive. I readily agree that Congress controls the appellate jurisdiction of the federal courts of appeals. Federal courts cannot hear appeals they have not been authorized to hear. But what they have been authorized to hear is appeals from “final judgments.” The sole reason for thinking this case was “not final” is that the panel decided to make it so. If the parties agreed to dismiss a claim that neither thought was sufficiently important to their case or controversy, that seems like an agreement that text, purpose, and common sense dictate the courts must honor. After all, given that the parties stipulated to the dismissal of the remaining claim, there was in fact nothing else for the district court to do but enter judgment, which is the textbook definition of a “final judgment.” I fear that shoring up the opinion with lofty invocations of our constitutional structure and the limited powers of Article III courts, while perhaps heartwarming and patriotic to all right-thinking members of the conservative legal movement, is nothing more than window dressing to poor legal reasoning.

The issue discussed here is, as I’ve noted, mundane. It is lawyer’s lawyer stuff. It involves the greasy machinery of day-to-day litigation. But the rules governing civil litigation in fact have a purpose, and it is one that is spelled out, and it is one that is supposed to bear on the meaning of the Rules when courts interpret them. It is barely explicable to me that the court did not consider it at all, but was so reflexive in deciding that “action” unambiguously refers to the entire lawsuit, failing to even discuss other subsections of the rule that cast serious doubt on that assertion, or to substantiate the claim that “action” is so limited in its meaning. My worry is that fealty to “isms” and “ists” are leading courts to short-cut interpretation of the law, rendering it counterproductive and even irrational (and remember, this is not a partisan problem; some say that “We’re all textualists now”). The outcome federal courts have landed on with respect to Rule 41(a)’s voluntary dismissal provision is directly at odds with the purpose of the federal rules—the just, speedy, and inexpensive resolution of civil cases. Quipping that such an outcome is required “because the text says so” strikes me more as fetishitic textualism, as the tail wagging the dog, than legal interpretation, and something we ought to try to get away from.