Linguistics & Language Neurology

The principle of the arbitrariness of the sign is not doubted by anyone, but it is often easier to discover a truth than to assign it its rightful place. This principle dominates all of linguistics — its consequences are innumerable. It is true that they do not appear all at once with equal clarity. It is after many detours that you discover them, and with them the primordial importance of the principle.

— Ferdinand de Saussure1

If you’re reading this, you probably understand English. It happens to be the de facto and sometimes de jure language of the U.S. legal systems,2 but some forms of “English” are more equal than others. Much work has reckoned with the contours of due process when people come to court with little to no English proficiency,3 but what about dialectal misinterpretation?4 English has numerous dialects, many birthed on what is now U.S. soil, but state and federal law have no coherent dialectal jurisprudence.5 The potential for error is worrying.6

Here, the goal is not to prove errors happen but to see whether the Constitution cares errors happen. It does. More precisely, the Due Process Clauses of the Constitution demand that the executive and judicial branches maintain procedures to avoid inaccurate transmission of linguistic data that adversely affects litigants. A mouthful to be sure, but the argument is that it is a violation of procedural due process to maintain procedures that will reliably cause misinterpretation of plain English and make it harder for litigants, especially criminal defendants, to win their cases. Intuitively, something’s amiss when the legal system, seemingly arbitrarily, messes up when interpreting some forms of English and not others. Past scholarship on dialect has brought invalu­able attention to the subject, and this Note seeks to continue that burgeoning tradition by showing how the status quo raises constitutional concerns and by serving as a resource and model for dialectal analysis going forward.

Part I showcases a few prominent English-to-English errors the legal system has made before, using Black English as the lens. One may wonder at the choice to use examples from only Black English when the point applies to any dialect. This is because of the insidiousness of the errors. Black English is a widespread dialect and one whose population of speakers is disproportionately represented in criminal adjudication, where colloquial testimony often features prominently. The point is general, but the readily available evidence is not. Furthermore, while it is probably true that racial minorities disproportionately bear the brunt of interpretive mistakes, any equal protection implications are for a different Note.7 Here, the focus is on language qua language, which, while correlated with race, needn’t be inextricably tied to it. Part II demonstrates that the legal system should think of these mistakes as procedural in nature with judicially administrable remedies. Part III argues that the Constitution has an open door for dialectal due process claims. And Part IV tours English legal history and documents the expansion of dialectal diversity to show how principles of linguistic fairness run deep.

The linguistic mistakes here are not the relatively predictable ones that arise from bad or no interpreters for a non-native English speaker. Instead, these mistakes happen when a native English speaker, indeed someone who might not speak any other language whatsoever, has difficulty being understood, which affects a litigant’s case. Misinterpretation can happen during live trial testimony or even before the police have begun an interrogation.8

These errors are both dangerous and insidious. They are dangerous because cases big and small are won and lost on the minutiae of language, and they are insidious for two reasons. First, if someone misinterprets speech and then carves that misinterpretation into the stone of the judicial record, it might be nigh impossible to uncover that a mistake happened at all. And second, if a monolingual anglophone judge from Macon, Georgia, hears Spanish, the judge knows they don’t understand, but if that same judge hears someone from England say “biscuit” and thinks of their grandmother’s delightful gravy-soaked masterpieces, the judge might not even realize their mistake.

The careful work of estimating exactly how much dialectal misinterpretation happens is for another day, but here, a few examples of the misinterpretation of Black English will serve to illustrate that it happens and matters at least sometimes. In each of these cases, a dialectal misinterpretation occurred that did indeed matter or obviously could have mattered.

Before embarking on this parade of misadventure, a note on something linguists have been screaming from the rooftops for decades, but maybe from rooftops a little too far out of the earshot of the legal system: No dialect is superior or more “correct” than another. Southern English, Black English, Chicano English, Appalachian English, American Indian English, or what have you are not degenerate, lazy, sloppy, or merely slang. Each dialect has an internal structure with rules.9 It’s possible to get things wrong.

What has become standard English in the United States is merely one dialect that had the historical fortune of being propelled to something approaching formal codification.10 It is certainly the written lingua franca, but faithful interpretation requires approaching the language on its own terms. English dialects vary immensely, and any attempt to make a consistent distinction between language and dialect is doomed from the outset.11 Linguists like to say that “a language is a dialect with an army and a navy”12 and that “there are as many languages as speakers.”13 With that in mind, consider these errors.

The “lawyer dog” case is probably the most famous recent clash between the legal system and English dialect. In 2015, New Orleans police wanted to interrogate a twenty-two-year-old black man named Warren Demesme on suspicion of sexual assault.14 Police had brought him in for questioning once before, and Demesme was reportedly getting frustrated, so he said: “if y’all, this is how I feel, if y’all think I did it, I know that I didn’t do it so why don’t you just give me a lawyer dog cause this is not what’s up.”15

The police did not give Demesme a lawyer, and he confessed.16 While Demesme was awaiting trial, his attorneys filed a motion to suppress the confession because the police got it out of him only after an unheeded invocation of the right to counsel.17 The prosecution argued that the statement, which the district attorney’s office provided as written above, was equivocal and therefore did not constitute a request for a lawyer.18 Eventually, the dispute got up to the Supreme Court of Louisiana, where the court denied Demesme’s petition.19

Justice Crichton additionally concurred. He argued that Demesme’s “ambiguous and equivocal reference to a ‘lawyer dog’ does not constitute an invocation of counsel that warrants termination of the interview.”20 He relied on Davis v. United States,21 where the U.S. Supreme Court held that the statement “[m]aybe I should talk to a lawyer” was ambiguous.22

To anyone scarcely familiar with Black English, it is painfully obvious that Demesme was using “dog” here (or, maybe had the defense lawyers been the ones to transcribe the statement, “dawg”) as a more familiar version of “sir.”23 He could have just as easily said “nigga,” “dude,” “man,” or “my guy.” If Demesme had said, “lawyer sir” or “lawyer man,” there would be little debate. This dialectal misinterpretation is likely clear to many who do not speak Black English, but as the next few examples will show, sneakier errors can happen as well.

In the following case, a federal judge in dissent misinterpreted the Black English present tense as possibly being the past tense, and the majority didn’t disagree. The admissibility of one piece of evidence, whether Joseph Arnold had a gun, hinged on whether a particular statement fit the excited utterance exception to the hearsay rule.24 The majority thought yes, the dissent no. The majority and the dissent also adopted different versions of the testimony in question from a Black woman25 named Tamica Gordon. The majority had her as saying: “I guess he’s fixing to shoot me.”26

Judge Moore disagreed and wrote that “[a]fter listening to the tape multiple times” she did not hear Gordon say “he’s fixing to shoot me” but instead “he finna shoot me.”27 The difference mattered because Judge Moore believed that “[t]he lack of an auxiliary verb renders determination of whether Gordon intended to imply the past or present tense an exercise in sheer guesswork.”28 And conditional on Gordon having said “he finna shoot me,” the majority did not disagree with the dissent’s linguistic analysis.29

But the linguistic analysis was wrong and the tense unambiguous. In Black English, the auxiliary verb in a sentence like this is omittable in the present tense, and only in present tense.30 So “he shooting” means “he is shooting,” not “he was shooting.” Insofar as the admissibility of the evidence turned on whether “he finna shoot me” was in the past tense, Judge Moore’s analysis was incorrect.

Finally, another disturbing feature of Judge Moore’s reasoning. To define “finna,” she used Urban Dictionary, arguing that the consensus nature of the site made it “unusually appropriate” for defining “slang, which is constantly evolving.”31 First, Black English, including “finna,” is not slang. And second, while in this case the definition she used was only marginally wrong32 and likely would not have affected her decision, Urban Dictionary is not the most reliable source. For the uninitiated, this is a user-generated definition site. To demonstrate why the use of Urban Dictionary is troubling, here is one definition of “judge” that is currently on the site:

The unmerciful uncivilized unfair pieces of shit they hire in the so-called “judicial system” which is about the biggest crock of shit there is out there.

Judges are pansies, often punishing the innocent and letting the guilty walk free. That’s why nobody has faith in the judicial system anymore. You’re better off to take the law into your own hands.

Ever been to divorce court? The judge almost always hears out the womans side of the case and totally ignores the mans side. . . .33

The madness continues, but for propriety’s sake the rest of the definition has been omitted. Despite these warning signs, Judge Moore is not alone in relying on Urban Dictionary for definitions.34

Judges don’t always have direct access to a recording like they did in the previous case. A recording might not exist at all, in which case it is up to the transcriber to ensure accurate transmission. Consider one jail call from 2015.

Two linguists listened to a recording of a call the police had transcribed. They noted two particularly important errors. When the suspect said, “He come tell (me) bout I’m gonna take the TV,” the police had transcribed “??? I’m gonna take the TV,” and where the suspect said “I’m fitna be admitted” the police had “I’m fit to be admitted.”35 If these transcripts got to a trial, they could make a dangerous difference.

This is just one phone call. In a now-landmark study, experimenters tested certified court reporters on Black English, and they failed in dramatic fashion:

Despite certification at or above 95% accuracy as required by the Pennsylvania Rules of Judicial Administration, the court reporters performed well below this level . . . . 40.5% of the utterances were incorrectly transcribed in some way. The best performance on the task was 77% accuracy, and the worst was 18% accuracy. . . . [T]he very best of these court reporters, all of whom are currently working in the Philadelphia courts, got one in every five sentences wrong on average, and the worst got more than four out of every five sentences wrong, under better-than-normal working conditions, with the sentence repeated.36

There are obvious limitations to this study.37 It is only one study with only twenty-seven court reporters from only the City of Philadelphia.38 But the potential danger of inaccurate transcription is clear.

Just as worryingly, transcribers sometimes intentionally change dialectal grammar in an effort to “sanitize[]” what they see as defects,39 transmogrifying meaning. Just like writing down a speech loses tone of voice, translating dialect might elide important information if the transcriber doesn’t know what to look for. For example, Black English has more aspectual markings than standard English. So, “he be running” is different than “he running.” The latter means “he’s running,” while the former means something like “he habitually runs, but not necessarily now.” A transcriber who doesn’t know that fact might think “be” is a mistake and transcribe “he is running,” changing the meaning.

If a dialectal speaker writes something themselves, no one can mishear or mistranscribe, but errors still happen. Cedric Antonio Wright, a defendant, responded “Yea” in a Facebook message to the question of if he was “tryna get ah glick.”40 The Eighth Circuit held that the “Facebook conversation revealed that Wright attempted to trade” for the gun.41 The panel’s ultimate conclusion that a reasonable jury could have determined Wright had the gun is correct given all the evidence, linguistic and not, in the case. However, the conclusion, insofar as the judges tried to make it, that answering in the affirmative to “tryna” reveals an “attempt” is false.

A lesser-known feature of Black English is the bivalence of “tryna.” It can mean “attempting to” as its etymology from “trying to” would suggest, but especially in questions or negative sentences, the term has another meaning — to desire. “You tryna eat?” in most contexts does not mean “are you attempting to eat?” but rather “do you want to eat?”42

Similarly, “you tryna get ah glick?” might not mean “are you attempting to get a Glock?” but rather “do you want to get a Glock?” That is, Wright could have, in theory, responded, “Yeah, I want and need one, but I can’t because that would be illegal.” Even in situations where judges have direct access to written records in the originator’s own hand, dialectal errors can still happen because judges are unaware of the differences.

The previous Part showed that errors can occur at any point in the process, whether professional transcribers are hired or judges have direct access to writing or audio. The goal of this Part, then, is to show that these are not just unfortunate, inevitable errors, but unfortunate, preventable errors — preventable through procedure. Consider these potential remedies.

To ensure reviewability and a complete record, judges, law enforcement interviewers, and transcribers should explicitly record the dialect they are dealing with as specifically as possible. It might be very difficult to tell if something is written in an unfamiliar dialect without reference to something besides the text itself,43 so making sure a police interview or a jail call that gets transcribed indicates the dialect spoken opens the door to more self-conscious interpretation. Suppose the transcript provided to Justice Crichton in the lawyer-dog case had “BLACK ENGLISH” written on the first page. Of course, misinterpretation can still happen, but in that situation, any prospective interpreter must acknowledge that if they are going to hang their hats on technicalities, they have to contend with dialect as well.

A second reason to record the dialect in question when transcribing or interpreting is future proofing. An expert at trial might find it useful to have an earlier assessment of the dialect in question, and an appellate court might be more likely to do a robust linguistic analysis if the dialect is apparent from the get-go. Without recording the dialect, during trial or on appeal, everyone might have to just guess at the proper interpretive framework. Besides making interpretation more self-conscious and reviewable, acknowledging the dialect serves to advance the legitimacy of nonstandard dialects and to promote popular knowledge that they exist and impact court proceedings.

Finally, this idea is not so crazy. Case law already implicitly acknowledges the dialects’ legal relevance. Courts have admitted lay testimony of accent to identify people in the tradition of “linguistic profiling.”44 The most famous example is probably the O.J. Simpson trial — Mr. Johnnie Cochran, Simpson’s lawyer, failed in his objection to the question: “When you heard that voice, you thought that was the voice of a young white male, didn’t you?”45 — but the Kentucky Supreme Court’s language is more explicit:

No one suggests that it [is improper for a lay witness] to identify [a voice] as female. We perceive no reason why a witness could not likewise identify a voice as being that of a particular race or nationality, so long as the witness is personally familiar with the general characteristics, accents, or speech patterns of the race or nationality in question . . . .46

Now, language is not intrinsically tied to race, as the Fifth Circuit has pointed out,47 but insofar as eyewitness testimony as to race is reliable enough to be admissible in a court of law (not to argue it should be), earwitness testimony as to race is astonishingly reliable. One experiment found that people are able to correctly distinguish between Black English speakers, Chicano English speakers, and Standard American English speakers more than seventy percent of the time — when they only heard the word “hello.”48 If courts acknowledge the existence of dialects when identifying suspects, it stands to reason they should acknowledge their existence when it comes to interpretation.

Pending further research on the best form for such instructions, judges could give juries cautionary instructions when dialects with public opprobrium show up in the courtroom. Factfinders might have prejudice against certain dialects. Linguists have shown that people are very good at identifying dialects very quickly49 and that potential jurors are prejudiced against certain dialects, notably Black English.50 Some find speakers of Black English to be less believable, less trustworthy, more criminal, less comprehensible, and more likely to be in a gang.51 Judges could explicitly instruct jurors that dialect, accent, and nonstan­dard grammar have no bearing on the truth of testimony or the person’s potential guilt. This might prompt self-conscious deliberation of those prejudices.

The actual instruction would need to be more specific and probably different depending on the testimony presented, but if a speaker of Black English were to testify, the judge might say the following to the jury before they do:

The next witness you will hear speaks Black English. This is a valid dialect of English and not wrong. Some things you hear might be more difficult to understand. Some things you hear might have a different meaning than you might initially think because of grammatical differences. But you must take care not to allow the differences in language alone to affect your judgments of the witness’s credibility. It would be unfair to the parties to ignore or discredit someone’s testimony just because of how they speak.

This Note strongly welcomes and encourages further-refined jury instructions that are most effective and avoid abuse. This particular instruction might very well not do that, and empirical scholarship might have something to say.

A word about implicit bias. While there is almost no reason to doubt implicit bias exists, it is important to note that there is little empirical reason to believe conventional implicit bias training works in changing behavior.52 Similarly, research on the comprehensibility of jury instructions and the effectiveness of limiting instructions and admonitions is far from promising.53 That said, cautionary instructions, which the above model attempts to be, have mixed results in the empirical literature, resulting in either no or a slightly positive effect.54 Given the jury instructions’ tiny cost, further research into benefits and abuses is extremely worthwhile.

The nuclear option is to get interpreters. This might sound crazy at first, but in 2010 the Drug Enforcement Administration released a memo to much media fanfare requesting nine “Ebonics” translators.55 The benefits of a competent translator are obvious: reduction of misinterpretation and formal recognition of the dialects as valid. And several prominent linguists think such an idea for Black English is at least going in the right direction.56

Now the drawbacks. First, no standardized tests exist for Black English and many other dialects, so for the time being, consistently mea­suring competency might be impossible. Second, interpreters are costly to train and hire. Third, having a native English speaker use an interpreter might be seen as a slight. And finally, if all there is is a written transcript, an interpreter might be of only limited value.57 One partial solution is to increase the number of jurors familiar with dialects important to the trial.58 When possible, this would help because they might act as informal interpreters,59 but in the case of less common dialects or if the trial happens in a place far from the epicenter of an important dialect, it might not be feasible.

Anyone interpreting dialect, especially judges, should use the best evidence available for determining the meaning of the language in question. While science is never perfect, it only makes sense that borrowing the tools that linguists, lexicographers, semioticians, and historians provide will make interpretation better.

The first step is to not use sources like Urban Dictionary if another option exists. Judges should strongly disfavor any resource that is publicly created, has little moderation, and does not cite sources. For dialects a judge speaks, they might be able to discern what is useful and what is not from less rigorous academic sources, but to rely on such sources for language the judge is unfamiliar with is dangerous.

The second step is to find reputable sources on the dialect in question. A somewhat more accurate, quick-and-dirty, crowdsourced definition site is Wiktionary, which at times explicitly notes the dialect of the word present, gives etymologies, and cites sources. If you question its usefulness, ask yourself if you know the Black English definition of “kitchen.”60 One of Wiktionary’s main limitations is its focus on words and not grammar. Furthermore, many dialects do not have formal dictionaries61 or textbooks, so the next best thing is academic linguistics sources. As with any specialized area, these papers might be opaque to anyone without specialized training, which is unfortunate, but the law asks judges to do many things other than pure legal interpretation.

The Federal Rules of Civil Procedure acknowledge this interdisci­plin­ary approach. Rule 44.1 allows judges to consider “any relevant material or source” when trying to figure out what foreign law means.62 And what this means when judges are faced with other languages can engender colic. Judges Posner and Easterbrook thought “[j]udges should use the best of the available sources,”63 which meant looking at official translations and secondary literature to determine foreign law instead of relying on party-provided expert testimony, which “adds an adversary’s spin.”64 Furthermore, Judge Posner had characteristically colorful words, arguing that the United States’s “linguistic provincialism does not excuse intellectual provincialism.”65 Similar reasoning applies to what judges should do when it comes to dialect. It seems equally provincial, if the term can be excused, to uncritically adopt the prosecution’s transcript when dialect is involved, as happened to Warren Demesme.

If possible, audiovisual or audio recording of statements would reduce error and single points of failure in the system. If someone mistranscribes, it’s permanent unless the source remains. Compared with mere acknowledgment and citing reputable sources, recording and storing audio have more-than-negligible costs,66 but they’re still probably much cheaper than professionally qualified court interpreters who get $495 per day in federal court.67 Recording’s main drawback is its inability to be universally applied. If the only thing the court has is earwitness testimony of something that happened out of court, nothing can be done. But when it is possible for the government to record testimony, doing so would both reduce error in the first instance and make it possible to remedy errors that do happen.

The shocking statistics from the Philadelphia experiment, if anywhere near generalizable, indicate that federal and state governments should mandate training in common dialects for all transcribers, and the National Court Reporters Association should create standard curricula for transcribing both common and uncommon dialects. Having stan­dards will both reduce error and make accurate transcription more accessible. Currently, at the federal level, the Judicial Conference recognizes as certified those who pass the Certified Realtime Reporter exam,68 which requires an accuracy of ninety-six percent on five minutes of real-time testimony at two hundred words per minute.69 Notably, the Association has no standardized testing, training, or certification for dialect.70 Building that infrastructure will be costly and take time, but the current certification process might mean very little for many English speakers.

Procedures that produce dialectal misinterpretation create constitutional concerns. Just because a particular procedural safeguard would reduce the likelihood of error does not mean the government has to do it, but courts do need to impose some measures. The question here is not whether the state is depriving someone of a right in the first place but whether dialectal misinterpretations implicate due process at all. They do for the simple reason that the “fundamental requirement of due process is the opportunity to be heard ‘at a meaningful time and in a meaningful manner.’”71 Many courts have required an interpreter when a litigant understands little to no English,72 but this Part argues intra-English issues are also worthy of remediation.

“For all its consequence, ‘due process’ has never been, and perhaps can never be, precisely defined.”73 But if there is a flagship article for what procedural due process means, it is Judge Friendly’s “Some Kind of Hearing,” and if there is a flagship case for what procedural due process means, it is Mathews v. Eldridge.74 This Part’s work is to analyze which factors dialectal misinterpretation implicates and why consistent misinterpretation can lead to less-than-meaningful hearings. As the title of Judge Friendly’s article hints, current law around what process is due is fluid and depends on the particular circumstances, and since dialectal misinterpretation transcends circumstance, the Constitution will demand no uniform rule. The conclusion here is not that such and such a procedure with respect to dialect is required but rather that some procedure with respect to dialect is required.

Mathews requires an opportunity to be heard “in a meaningful manner.”75 Reasonable minds can disagree about what this means, but analyzing Mathews’s balancing test shows that dialectal misinterpretation implicates exactly what judges must consider when shaping the contours of constitutionally required process. Mathews commands that to decide whether the Constitution requires more or different process in a case, courts must consider:

First, the private interest that will be affected by the official action; second, the risk of an erroneous deprivation of such interest through the procedures used, and the probable value, if any, of additional or substitute procedural safeguards; and finally, the Government’s interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedure requirement would entail.76

If the mistakes showcased in Part I tell anything, it is that these mistakes do matter at least sometimes, and in big ways. This section will touch on all three factors and the case’s axiomatic proclamation that due process requires an opportunity to be heard in a “meaningful manner.”

Dialectal misinterpretation can feature in any case, so the private interest will vary considerably. Empirical work is necessary to determine when dialect does indeed come up most prominently, but it will certainly feature in cases where litigants have a lot at stake, whether that’s criminal charges like in State v. Demesme77 and United States v. Arnold,78 deportation, or parental rights termination. Dialect is always a potential problem.

Moving to the second factor and the risk of erroneous deprivation, more work is necessary before anyone can in good faith make an estimate as to exactly how much dialectal misinterpretation happens and matters for the deprivation. But if the Philadelphia experiment proves generalizable, mistakes might be happening with alarming frequency. The probability will naturally differ by dialect, region, and court. And, when one considers how badly certified court transcribers did in the study that does exist, it’s a hard sell that the risk of erroneous deprivation from the lack of a coherent approach to dialects is too small to be judicially cognizable. For one, Louisiana should have thrown out Demesme’s confession.

And, as Part II demonstrated, there do exist procedural safeguards that have a good chance of reliably protecting against dialectal misinterpretation, both in the short and the long term. If the Louisiana Supreme Court had acknowledged that Demesme was speaking Black English and looked to academic sources on the dialect, he probably would have prevailed. If police-employed jailhouse transcribers had training and testing on dialects, they would likely not transcribe “fitna be admitted” as “fit to be admitted.” Getting precise statistical mea­surements of how often these problems occur is not possible at the moment, nor is knowing exactly how much any particular procedure will help. That said, Mathews itself said that “[b]are statistics rarely provide a satisfactory measure of the fairness of a decisionmaking process.”79 When it comes to misinterpreting dialect, the principal unfairness is that by having the misfortune of speaking or relying on testimony in the wrong kind of English, defendants find a legal system unprepared to treat them fairly.

On to the government’s interests. Every procedural safeguard will have a different cost. Interpreters are likely the costliest, followed by dialect certification for transcribers. Acknowledging dialects, instructing juries, and using reputable sources are basically costless. And increasing the use of audio recording lies somewhere in between. As with the entire analysis, the strength of the government’s interest is context dependent. But it seems unlikely that the cheapest procedures’ minimal costs would often overcome the specter of erroneous deprivation. Audio recording already happens in many situations, and transcribers already exist. Increasing the adoption of audio recording and improving the accuracy of transcribers for dialects would impose real costs, but for important deprivations, such as physical liberty, due process might often require both. The argument for interpreters is weakest, particularly if there are ways to improve the accuracy of transcribers, but it is not implausible that they should be required for tricky dialects when physical liberty is at stake.

Finally, as has been said but bears repeating, linguistic issues go to due process’s root because the requirement of a “meaningful” opportunity to be heard requires, if anything at all, the hearer to understand the litigants. Dialectal barriers infringe upon understanding at the most basic level, even when the interpreters might not think so. In some circumstances, the litigant or their counsel might catch mistakes as they happen, but in the course of litigation there is not constant confirmation of a sort of consensus ad idem. So, the system cannot place the duty of clearing up misinterpretations at the feet of litigants.

Judge Friendly’s article is incredibly influential, cited by more than 300 cases and by the Supreme Court itself eleven times.80 The article lists eleven factors “that have been considered to be elements of a fair hearing, roughly in order of priority.”81 The factors are not a Restatement-like list of what proceedings require, but a framework for the ways a procedure might implicate due process.82 Dialectal misinterpretation strongly implicates several of these factors, including the bias of the tribunal, a decision based only on the evidence presented, and the making of a record. More tenuously, it also implicates an opportunity to present reasons why the proposed action should not be taken.

Starting with the most fundamental, a tribunal that does not prepare itself for dialectal misinterpretation is biased. Judge Friendly called an unbiased tribunal “a necessary element in every case where a hearing is required.”83 And unselfconscious dialectal interpretation can bias a tribunal. Juries are more likely to evaluate speakers of certain dialects poorly,84 and that is clearly a thumb on the scale against those who rely on such speakers’ credibility. More generally, even if the misinterpretation is not the result of invidious evaluation of the speaker, a tribunal that fails to treat each dialect on its own terms and instead forces them all to conform to one mainstream structure has biased itself against speakers of all those dialects it fails to recognize. Dialect A speakers get good interpretation, but Dialect B speakers get bad interpretation and are thus less able to press their case to the court.

Next, consider the idea that when judges or juries freestyle interpretations of dialectal testimony or base their interpretations on unreputable sources of linguistic information, they are making a decision based on evidence other than that presented. It is not just a mistake but also an injection of unnecessary randomness into the decisionmaking process. Whether the interpreter knows it or not, if they are unprepared for dialects, they might, based on the random similarities or differences with their own dialect, hold incorrectly. The fact that “dawg” sounds like “dog” is historical happenstance, but the result is that Louisiana deprived someone of the right to counsel. An illustration of the point in the extreme: a judge, knowing a smidge of Spanish, would be mistaken if they thought a person who said they couldn’t come to court because they were “embarazada” meant they were too embarrassed to come in. Similarly, a judge cannot conclude “tryna” means “attempt” simply because it sounds like the Standard English “trying to.” Otherwise, stochastics, not evidence, is determining outcomes.

Third, consider the record. The record’s importance is so ingrained that Judge Friendly felt “Americans are as addicted to transcripts as they have become to television; the sheer problem of warehousing these mountains of paper must rival that of storing atomic wastes.”85 And the main purpose of this mass of paper, so the argument goes, is the ability for judicial review or administrative appeal.86 When stenographers and transcribers are systematically bad at interpreting English, then, they vitiate that purpose. The Supreme Court said “denial [of free transcription services to indigent criminal defendants] is a misfit in a country dedicated to affording equal justice to all and special privileges to none in the administration of its criminal law”87 because “[t]here is no meaningful distinction between a rule which would deny the poor the right to defend themselves in a trial court and one which effectively denies the poor an adequate appellate review accorded to all who have money enough to pay the costs in advance.”88 As has been shown, dialectal misinterpretation can not only reduce a transcript’s accuracy but also introduce harmful errors, like changing hearsay (“He come tell (me) bout I’m gonna steal the TV.”) to a confession (“??? I’m gonna steal the TV.”).

Finally, dialectal difficulties implicate the ability to present reasons why a proposed action should not be taken. Pro se litigants might have an “inability to speak effectively for” themselves,89 not only because they don’t understand the law but also because their dialect might not reach the judge as easily. In the situation where a litigant represents themself, if they speak a dialect for which the hearer is not prepared, how much of an opportunity is it?

Principles of linguistic justice run deep in English legal tradition. And, maybe counterintuitively, dialectal due process was likely much less of a problem in the past because justice was more local and because English probably has more varieties today than ever. So, attempts to foreclose dialectal due process claims on the basis that they didn’t exist historically (assuming they didn’t arguendo) are misguided.

The “primary guide in determining whether the principle in question is fundamental is, of course, historical practice.”90 This idea has special weight when considering the state power to regulate procedure because:

[I]t is normally “within the power of the State to regulate procedures under which its laws are carried out,” . . . and its decision in this regard is not subject to proscription under the Due Process Clause unless “it offends some principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental.”91

As the previous Part shows, dialectal due process implicates fundamental principles like unbiased tribunals, but historical practice and circumstances also show linguistic fairness has a pedigree in its own right. The aim is to show not that earlier courts self-consciously accommodated dialect in particular but that they were structurally less likely to have dialectal-misinterpretation issues because juries worked radically differently and there were simply fewer dialects of English to contend with — a problem that will likely only get worse. This Part highlights trends reaching back to pre-Norman England and uses principles of historical linguistics to argue that early courts didn’t necessarily need self-conscious dialectal due process, but courts today do.

The idea of linguistic due process is old. The Pleading in English Act 136292 rebuffed the “great Mischiefs” resulting from the fact that people had “no Knowledge nor Understanding of that which is said for them or against them” in court, which was in French.93 The statute felt that “reasonably the . . . Laws and Customs [the rather shall be perceived] and known, and better understood in the Tongue used in the said Realm.”94

Fourteenth-century statutes likely don’t explicitly declare dialect’s importance,95 but that does not foreclose constitutional dialectal due process because the Constitution incorporates many common law ideas of fairness. The Act was the beginning of an almost seven-century-long tradition of conducting court in the vernacular. To say, then, that the idea of linguistic misinterpretation (and one subset of that, dialectal misinterpretation) has no historical basis is false.

Even if no dedicated procedures for dialectal due process existed historically, justice’s local nature meant that procedure had dialectal protection baked in. If the lion’s share of people investigating and passing judgment on you are your literal neighbors, they are much more likely to understand or even speak your dialect. Local justice has been significantly diluted from the prebiotic broth of medieval England.

First, terminology. A “hundred” was the second-smallest administrative unit in England, bigger than a parish but smaller than a shire.96 The name comes from its origins as consisting of one hundred “hides,” a maddeningly inconsistent unit thought equivalent to the land needed to support a peasant family.97

Hundred courts empaneled juries from the hundred.98 And in Anglo-Saxon England, the most important interaction people had with the Crown occurred in these courts.99 To be sure, some historical Germanic practices, such as defending court judgments by duel,100 fell out of favor, but many of the practices and procedures developed in hundred courts became a model “all over England in the courts of the manors.”101

The earliest juries were something like “a body of neighbours . . . summoned . . . to give upon oath a true answer to some question.”102 The key word is “neighbours.” Hundred jurors would be coming from a much smaller pool than juries today in the United States, meaning they were much more likely to truly be from the same community and speak the same dialect. The number of hundreds into the nineteenth century was somewhere around 800.103 The English population from 1790 to 1800 was around eight million.104 That means a pool of 10,000 people per hundred. For reference, Harvard University has a workforce of around 13,000 people with a student population of around 23,000.105 The low population density itself was a procedural safeguard against dialectal misinterpretation because the factfinder was more likely to speak the relevant dialects.

The average judicial-district population density in the United States at both state and federal levels falls well outside the average in England around 1800 before the nineteenth-century population explosion. The states average a judicial-district population density of around 16 times higher and the federal government’s ninety-four districts leave a popula­tion density 350 times higher.106 In sum, jurors today come from more populous districts and serve at random,107 meaning they are less likely to be from truly the same linguistic community as the litigants and witnesses.

Through the centuries these hundred courts slowly lost importance. By the 1830s only a few remained, with even fewer active,108 and other courts certainly existed and empaneled juries.109 Some had even smaller jury pools than the hundred courts.110 And during the hundred courts’ twilight in the 1800s, England saw the rise of county-level justice with justices of the peace or magistrates,111 who still make up about eighty-five percent of the English judiciary,112 and the county courts, which ascended in importance in the mid-1800s.113

But the structure protecting dialectal understanding was not immediately lost. The jurors’ level of involvement ensured a level of dialectal understanding. In the earliest times, jurors served because they knew facts concerning the case and the accused.114 These early juries were self-informing, investigating the facts separate and apart from the trial.115 And although jurors were no longer selected with input from the judges starting in 1730116 or self-informing, they were much more involved in the trial process than they are today. Jurors could ask their own questions, request more witnesses, and, crucially, volunteer their own pertinent knowledge about local custom, people, and places.117 Blackstone gave a categorical answer on jury involvement as it existed when he published the first edition of Book III of his Commentaries in 1768. He wrote that “the practice . . . now universally obtains, that if a juror knows any thing of the matter in issue, he may be sworn as a witness, and give his evidence publicly in court.”118 Such a notion is almost anathema today. But the lawyerization of the courtroom and the separation between the juror’s judicial and personal role was only beginning in this era.119 Since jurors were more local and involved in trial process historically, they would have had many more chances to clear up dialectal misinterpretation in view of the judge and litigants.

The analysis thus far makes a key assumption — that the number of dialects that exist has remained constant. I