Justice Young delivered the opinion of the Court.

Sarah N. Wariner, Audra Leigh Braden, Frisco, Jadd Fitzgerald Masso, Dallas, for Petitioner Dorchester Minerals, L.P.

Phil Johnson, Marc S. Tabolsky, Mark W. McBrayer, Lubbock, William Chris Boyer, Andrew Raymond Seger, for Petitioners Pam Stribling and John V. Price as Heirs and Successors of Interest to Noble H. Price, Key, Roger A., The Huffington Foundation, Bishop-Windham Family Limited Partnership, Key, Terry S., Davis, Stephen L., Keys, Kay Elaine, Davis, Arthur B., Stuckert, Jill Marie, George Dan Mulkey and Thomas J Mulkey, Trustees of the Mulkey Family Mineral Trust.

Kelley Clark Morris, Gabe Vick III, James J. Ormiston, Houston, for Petitioners PXP Producing Company, LLC f/k/a North Central Oil Corp. et al., Freeport-McMoRan Oil & Gas LLC.

Phil Johnson, Mark W. McBrayer, Lubbock, Marc S. Tabolsky, John L. Lancaster III, Dallas, William Chris Boyer, for Petitioners Lancaster, Jane R., Ryedale, LLC.

Frederick B. Wulff, Kevin P. Burke, James A. Holmes, Henderson, for Respondent McCullar et al., Dave Michael.

Marc S. Tabolsky, Phil Johnson, William Chris Boyer, Andrew Raymond Seger, Mark W. McBrayer, Spenser Piercy, Lubbock, for Petitioner Van Dyke, Susan Davis.

Don M. Connally, Abilene, for Respondent John W. Daniels and Wilma F. Daniels.

Forrest Jacob Wynn, for Other interested party Baylor College of Medicine.

Matthew David Stayton, Derek Lee Montgomery, Harold L. Hensley Jr., Midland, David E. Keltner, Fort Worth, P. Taylor Spalla, for Respondent The Navigator Group et al.

Judith A. Meyer, Houston, William Chris Boyer, Marc S. Tabolsky, Linda Broocks, Houston, Mark W. McBrayer, Lubbock, Phil Johnson, for Petitioner Trust, N.A., Raymond James.

Alan A. Aaron, Midland, for Other interested party Simmons et al., Warren R.

William Chris Boyer, Marc S. Tabolsky, Phil Johnson, Mark W. McBrayer, Lubbock, Willis E. Gresham Jr., Lamesa, for Petitioners Culley Ingram, Trustee of the CI Grandchild's Trust, Kerry Kantman, Trustee of the KK Grandchild's Trust, Alexander, Deborah L., Amanda Kay Livingston, Trustee of the AKL Child's Trust, McKenzie Ciliberto, Trustee of the MC Grandchild's Trust.

Ken K. Slavin, El Paso, W. Clayton Gaston, Midland, for Petitioners Deutsche Bank Trust Company, NA, and Irving Sitnick, Trustees of the Lucy G. Moses 12/22/58 Trust, Deutsche Bank Trust Company, NA, Trustee of the William L. Hernstadt 1937 Trust, Deutsche Bank Trust Company, NA, and William H. Hernstadt, Trustees of the William H. Hernstadt Estate Trust, Deutsche Bank Trust Company, NA, Trustee of the Henry & Lucy Moses Foundation Trusts.

Marc S. Tabolsky, William Chris Boyer, Ted M. Kerr, Mark W. McBrayer, Lubbock, Phil Johnson, for Petitioner Bridgewater, Jr., Preston.

Richard E. Booth, Midland, for Other interested party Endeavor Energy Company et al.

John T. Ferguson, Big Spring, for Other interested party Blissard, Bennie.

Robert M. Park, for Amicus Curiae Texas Land and Mineral Owners Association.

Phil Johnson, Timothy J. Henderson, Mark W. McBrayer, Lubbock, Marc S. Tabolsky, Andrew Raymond Seger, William Chris Boyer, for Petitioner Huttner, Sheryl Ann.

Christopher Michael Hogan, for Other interested party Powder River Mineral Partners, LLC, successor to Petrohawk Energy Corporation.

Brandon Archer, for Other interested party Williams et al., Elizabeth Rankin.

Ryan S. Mindell, for Respondent G.R. White Charitable Trust.

Jefery F. Carson, Sugar Land, Mark W. McBrayer, Lubbock, Marc S. Tabolsky, Phil Johnson, William Chris Boyer, for Petitioner McClellan, James G.

Kevin P. Burke, for Respondent Mills et al., Loretta.

James Jackson Napper, Lamesa, for Other interested party Chalmers, James D.

Michael T. Morgan, Midland, for Other interested party Huffington, Roy.

Alan Dane Waterman, Pro Se.

H. Alan Carmichael, Sweetwater, for Other interested party Mallard Royalty Partners d/b/a Texas Land Royalty, Ltd. et al.

Stephanie Lee, David W. Lauritzen, Midland, W. Bruce Williams, Emily Brown, Houston, Matthew Coolbaugh, Midland, for Respondent The Chase Trustees.

Christopher Kulander, for Amicus Curiae Kulander, Christopher S.

George Henry Rau III, Houston, for Other interested party Black Stone Minerals, LP.

Phil Johnson, William Chris Boyer, Marc S. Tabolsky, Mark W. McBrayer, Lubbock, for Petitioner Key, Howard W.

Marc S. Tabolsky, Phil Johnson, William Chris Boyer, Mark W. McBrayer, Lubbock, Judith A. Meyer, Houston, for Petitioner William Marsh Rice University.

Marc S. Tabolsky, Mark W. McBrayer, Lubbock, Andrew Raymond Seger, Ken K. Slavin, El Paso, William Chris Boyer, Phil Johnson, Scott Kendall, Dallas, for Petitioner Boyd Enterprises, Inc.

Justice Young delivered the opinion of the Court. Only in a legal text could the formula "one-half of one-eighth" mean anything other than one-sixteenth. But in the law, "one-half of one-eighth" sometimes equals one-half—in the context of reservations of mineral interests. Likewise, the law sometimes calculates one-half of 1,000 to be 600, not 500—in the context of contracts for rabbits. Those results may seem bizarre, unsatisfying, and literally fuzzy math. They can also be inefficient; resolutely adhering to the rules of arithmetic would more rapidly end litigation. The rules that courts must apply, however, are not primarily those of arithmetic but of textual construction. The rules of construction, in turn, reflect the principle that legal texts—including private-law documents like contracts, deeds, and wills—still bear the meaning that their words had when they were drafted, even if the use of the same words today might generate a different meaning.

This case involves the first seeming oddity mentioned above: the so-called "double-fraction" dilemma from antique mineral conveyances in which the parties insisted on using two fractions. We must stretch back nearly a century to determine the meaning of a 1924 deed's mineral reservation of "one-half of one-eighth." This is not our first case involving double fractions, and it is likely not our last. But building on our precedents, and focused on our duty to faithfully interpret any legal text, we anticipate at least substantially reducing the frequency of disputes about double fractions. We conclude that an accurate construction of the 1924 text requires us to accept that the equation "one-half of one-eighth" equals one-half of the mineral estate. Even if this were not so, nearly a century of the parties’ unbroken understandings and representations would require us to recognize that allocation of present-day ownership by applying the presumed-grant doctrine. We accordingly reverse the judgment of the court of appeals and remand the case to the trial court for further proceedings.

I

In 1924, George H. Mulkey and Frances E. Mulkey conveyed their ranch and the underlying minerals to G. R. White and G. W. Tom (who had a general partnership called "White and Tom") with the following reservation:

It is understood and agreed that one-half of one-eighth of all minerals and mineral rights in said land are reserved in grantors, Geo. H. Mulkey and Frances E. Mulkey, and are not conveyed herein.

After the deed's execution, both parties, their assignees, and various third parties engaged in numerous transactions and filings reflecting that each side of the original conveyance had an equal 1/2 interest in the minerals. This included further conveyances, leases, ratifications, division orders, contracts, probate inventories, stipulations, and various other recorded documents. In 1946, however, Ethel Stuckert, the Mulkeys’ daughter, wrote the following to her brother in a letter:

After several weeks of consultations ... between Mr. G. R. White, George [F. Mulkey] and [Young] J. [Mulkey], a contract was entered into whereby Mama, and Papa's heirs, will receive one half of the mineral rights on the old ranch land.

The record does not otherwise confirm that such a "contract" was in fact executed. If it was, it has been lost to time; at least, the record before us does not include it. Nor does the record explain why the parties would have wanted to enter into such a contract, whether any consideration was paid, or whether (if in fact executed) it was genuinely a "contract" at all rather than a clarification.

Nonetheless, both before and after that letter, one important feature remained constant. Specifically, for nearly ninety years after the original deed's execution, the parties (including new owners who received various interests) continued without exception to engage in transactions and to make representations about their ownership interests that were consistent with the understanding that each original side had always had a 1/2 interest in the minerals.

The present litigation represents the breakdown of that mutual understanding of equal ownership. The two sides are called either "the White parties" (those whose interests derive from White and Tom, the grantees, and who are respondents in this Court) or "the Mulkey parties" (those whose interests derive from the grantors, and who are petitioners here). In 2013, the White parties brought this trespass-to-try-title action after Endeavor Energy began to pay royalties from its drilling operation to both parties in equal shares. At stake is at least $44 million in accumulated disputed royalties.

The ownership of those (and presumably future) royalties turns on which side correctly interprets the deed's mineral reservation of "one-half of one-eighth." The White parties assert that the double fractions are merely an elementary arithmetic formula with no additional meaning, so that only a 1/16 interest was ever reserved. The Mulkey parties contend that the double fraction reflects a term of art common at the time the deed was drafted and that the use of this term of art reserved 1/2 of the mineral interest. Alternatively, the Mulkey parties assert that even if the deed had only reserved a 1/16 interest, they gained title to the remaining 7/16 by operation of the presumed-grant doctrine at some point after 1924 but long before the $44 million accrued.

After both parties filed various competing summary-judgment motions, the trial court entered an order granting the White parties’ motion for partial summary judgment on the construction of the 1924 deed. The order declared that the deed's reservation of "one-half of one-eighth of all minerals and mineral rights" unambiguously reserved only a 1/16 interest in the mineral estate. The trial court then denied the Mulkey parties’ alternative motion for partial summary judgment, which asked the trial court to declare that they owned title to 1/2 of the mineral interest by virtue of the presumed-grant doctrine or various other defenses. The court granted the White parties’ accompanying no-evidence motion for summary judgment with respect to the Mulkey parties’ "affirmative claims and/or defenses based upon theories of adverse possession, estoppel (both equitable and judicial), waiver, laches, estate misconception, presumed grant, lost deed, fraud, and failure to mitigate damages." All other claims were severed, and two other appeals involving the 1924 deed have been abated pending our decision in this case.

The court of appeals affirmed. 647 S.W.3d 901, 913 (Tex. App.—Eastland 2020). It held that the deed unambiguously conveyed 15/16 of the mineral estate. Id. at 908. The court concluded that the estate-misconception theory—the theory that the Mulkey parties press to justify their counter-arithmetical reading—had no role to play because the deed did not contain any conflicting provisions requiring harmonization and because the subject property was not burdened by a lease at the time of conveyance (or before then). Id. at 907–08. The court thus applied standard multiplication to determine the quantum of mineral interest reserved. Id. at 908. With respect to the Mulkey parties’ alternative claim, the court held that the presumed-grant doctrine did not apply, primarily because there was no "gap" in the chain of title. Id. at 909.

II

We hold that the trial court and the court of appeals erred in holding that the Mulkey parties do not have a 1/2 interest in the minerals. Two distinct paths—the construction of the original deed and the presumed-grant doctrine—lead us to the same conclusion. Not all cases will require analysis of each path, so we address both to explain how they differ and to identify their respective requirements.

First, we conclude that the deed itself reserved a 1/2 interest in the mineral estate. Antiquated instruments that use 1/8 within a double fraction raise a presumption that 1/8 was used as a term of art to refer to the "mineral estate." That presumption is readily rebuttable, however. If the text itself has provisions—whether express or structural—illustrating that a double fraction was in fact used as nothing more than a double fraction, the presumption will be rebutted. But the presumption is not rebutted here. Nothing in the text of this deed suggests that rote multiplication was intended, and it is not inconsistent with any part of the deed to read 1/8 as a term of art that references the entire mineral estate. Even if we assumed for argument's sake that this was wrong, though, the presumed-grant doctrine would confirm that the Mulkey parties today have title to 1/2 of the mineral estate. We address these points in turn.

A

We begin by explaining how we construe the deed.

1

Whether the 1924 deed reserved a 1/2 mineral interest for the Mulkey parties or a 1/16 interest—or anything else—reduces to a question of textual interpretation. The fact pattern may seem odd to those not steeped in Texas oil-and-gas law, but the legal framework for analyzing this text is the same as for any other.

Unless otherwise defined in the text, courts will adopt a term's ordinary meaning. URI, Inc. v. Kleberg County , 543 S.W.3d 755, 763–64 (Tex. 2018). One fundamental premise, however, is that a text retains the same meaning today that it had when it was drafted. Thus, the ordinary meaning at the time of drafting remains the meaning to which courts must later adhere. We have made this basic point repeatedly, even in the very double-fraction context that we confront today: "Words must be given the meaning they had when the text was adopted." Hysaw v. Dawkins , 483 S.W.3d 1, 13 (Tex. 2016) (quoting Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 78 (2012)).

The meaning of an unamended text, in other words, is unaffected by the passage of time, linguistic developments, or the evolution of usage. These phenomena may affect our language by giving new meanings to (or subtracting old meanings from) any given word or phrase. But the original text does not evolve with the broader language. The test is what the text reasonably meant to an ordinary speaker of the language who would have understood the original text in its context. Whatever that meaning was then remains the meaning today.

For example, early Texas jurisprudence recognized that a contract for a "thousand" rabbits was understood to mean 1,200; reference to a "day" could mean ten hours in context. Dwyer v. City of Brenham , 70 Tex. 30, 7 S.W. 598, 599 (1888). Texts of that era using those terms still bear those meanings. The way to change a text's meaning is to change the text , not to observe that an unchanged text includes language that, unbeknownst to those who committed the text to writing, would at some point in the future eventually carry a different meaning.

The U.S. Supreme Court's decision in New Prime Inc. v. Oliveira , ––– U.S. ––––, 139 S. Ct. 532, 202 L.Ed.2d 536 (2019), provides a recent illustration. The Court there construed the term "contracts of employment" as used in the Federal Arbitration Act, which Congress enacted in 1925. Today, "contracts of employment" is a term of art that distinguishes those in formal employment relationships from independent contractors. Id. at 539. But the statute's meaning turns on what that phrase meant in 1925 , not what it would mean today. The Court thus undertook careful analysis of that phrase's public meaning at the time of the statutory enactment. It examined contemporary legal and lay dictionaries, judicial decisions, federal and state statutes, and the like. Id. at 540–41. The Court agreed that "modern intuition isn't easily squared with evidence of the term's meaning at the time of the Act's adoption in 1925," when the term "meant nothing more than an agreement to perform work." Id. at 539. In other words, "employment" was not used as a term of art in 1925. It broadly meant work done regardless of the relationship between the one doing the work and the one paying for it. That same meaning must govern the Act today.

Thus, our analysis does not turn on what we might think "one-half of one-eighth" would mean if written today. It does not matter whether that phrase would clearly mean 1/16, or whether the now-unusual step of spelling out a double fraction would make its meaning inherently ambiguous, or something else. Indeed, the challenge is not particularly legal in nature. Instead, it is to overcome the cognitive dissonance that arises because, at least at first glance, "one-half of one-eighth" seems unusually clear yet is alleged to mean something radically different from what we might expect. After all, it is certainly true that one-half times one-eighth did equal one-sixteenth in 1924 and at every other time in history. But 1,000 has always meant 1,000 (not 1,200) throughout history, too, even for rabbits; days have always had 24 hours, not 10. Setting aside our preconceptions and our instinctive resort to basic math, our only question is whether, in the context of a mineral-conveyance instrument from 1924, the double fraction reasonably referenced unchanging arithmetic at all. The analytical framework here—simply determining what a term meant and thus means—turns out to be exactly the same as it always is.

Precisely because we are addressing the ordinary process of ascertaining a text's meaning, we emphasize that the initial analysis remains confined to the four corners of the document as usual. We said that, too, in a case involving the double-fraction problem. See Garrett v. Dils Co. , 157 Tex. 92, 299 S.W.2d 904, 906 (1957). Our interpretation of any contract or deed primarily concerns the parties’ intended meaning. See, e.g., Myers v. Gulf Coast Mins. Mgmt. Corp. , 361 S.W.2d 193, 197 (Tex. 1962). But as always, we determine intent objectively by giving words their fair meaning. Matagorda Cnty. Hosp. Dist. v. Burwell , 189 S.W.3d 738, 740 (Tex. 2006). We do not start with extrinsic evidence, nor do we credit claims made in litigation of a secret or bespoke meaning that no one not privy to the code reasonably would have understood. True, "parties may freely define an ordinary word to have an unusual meaning; when they do, they rebut the presumption of ordinary usage. Without any textually expressed bespoke meaning, however, courts will adopt the ordinary usage as a matter of law." Perthuis v. Baylor Miraca Genetics Lab'ys, LLC , 645 S.W.3d 228, 236 (Tex. 2022) (footnote omitted).

The question, then, is what the ordinary usage of the textually undefined term "one-half of one-eighth" actually was. In other words, we do not begin by asking what the original White and Mulkey parties secretly or unusually might have meant by it. Merely indulging that question at this stage of the analysis would invite a flood of unreliable and destabilizing extrinsic evidence, which would be contrary to our precedents and hostile to the rights of parties to have their documents’ meaning measured objectively without being undermined by a later (perhaps much later) subjective inquiry.

Deeds provide a good example of why we insist on language bearing its ordinary meaning. Recording deeds and similar instruments is purposefully a public enterprise designed to elicit public reliance. "[T]he reliability of record title contributes mightily to the predictability of property ownership that is so indispensable to our legal and economic systems." Cosgrove v. Cade , 468 S.W.3d 32, 34 (Tex. 2015). A properly recorded deed, like the one at issue here, "provides all persons, including the grantor, with notice of the deed's contents," id. , which would be far less valuable without a consistent and stable judicial construction of terms used in deeds. The meaning of a deed, in other words, matters to the public writ large, not merely to those who wrote it. So important is it that these records are public and permanent that we recently overturned a decades-old default judgment foreclosing a tax lien largely because of the failure "to consult public deed and tax records," which would have revealed information necessary to achieve proper service on a defendant. Mitchell v. MAP Res., Inc. , 649 S.W.3d 180, 191 (Tex. 2022).

We therefore ask, as the Court in New Prime did, whether there is some objective reason that the double fraction in the deed at issue meant something other than its arithmetical result. URI , 543 S.W.3d at 765. Said differently, using extrinsic evidence of subjective intent to inform the language that the specific parties used would impermissibly trespass beyond the document's four corners. Courts frequently consult contemporary dictionaries because they convey objective and generally available—not subjective or bespoke—guides to meaning. Specialized or technical dictionaries can provide the same assistance for texts that arise in specialized or technical contexts. Indeed, the types of generally available sources that the Court used in New Prime to confirm that the meaning of "contracts of employment" has changed over time can benefit textual analyses of all kinds. The goal of such an inquiry is always to determine what a text could reasonably have meant to an informed but disinterested speaker at the time the text was written.

Determining what a text would have meant to a disinterested audience is an inquiry that is designed to confine courts to the four corners of the document and is a proper part of interpretation. See, e.g., U.S. Shale Energy II, LLC v. Laborde Props., L.P. , 551 S.W.3d 148, 152 (Tex. 2018) ("We may consider such circumstances to the extent they ‘inform, rather than vary from or contradict, the [instrument's] text.’ " (quoting URI , 543 S.W.3d at 767 )). As we discuss below, only if the text remains incapable of a clear meaning—and thus is unavoidably ambiguous—would we then move beyond the traditional, neutral, and objective tools of textual analysis.

2

This brings us to the now-familiar observation that, at the time the parties executed this deed, "1/8" was widely used as a term of art to refer to the total mineral estate. Notably, it is that fraction—not 1/3, 2/7, 6/241, or any other—that is so repeatedly deployed. Happily, we need not speculate as to why. Hysaw recently undertook the core analysis on which we rely today.

The Court there examined what it called the "Double-Fraction Dilemma" in the context of a 1947 will-construction dispute where a provision of the will bequeathed each child a "one-third of one-eighth royalty." Hysaw , 483 S.W.3d at 4. While recognizing that "[d]iscerning the nature of a particular royalty interest may be a simple matter when an instrument consistently uses single fractions to describe the interest," double fractions can present serious complications. Id. at 9. The Court looked to the objective circumstances as a necessary part of defining the terms in their context. Id. at 9–15. The Court identified two related circumstances that explain why 1/8 in such instruments did not typically bear its arithmetical meaning: the historical use of 1/8 as the standard royalty and the estate-misconception theory. Id. at 8.

Because these historical features confirmed that 1/8 was a widely used term of art, the Court expressly rejected "a mechanical approach requiring rote multiplication of double fractions whenever they exist." Id. at 4. Instead of simple multiplication, we held that "[i]ntent must be determined by a careful and detailed examination of the document in its entirety, rather than by application of mechanical rules of construction that offer certainty at the expense of effectuating intent." Id. at 16. Applying this analysis, the Court found that the term "one-third of one-eighth royalty" did not intend to relay a fixed 1/24 royalty, but instead a floating 1/3 interest in the royalty. Id. at 5.

Hysaw provides the foundation for resolving this case and many others. Nonetheless, it remains true that "[t]he proper construction of instruments containing double-fraction language is a dilemma of increasing concern in the oil and gas industry, as uncertainty abounds, disputes proliferate, and courts have seemingly varied in their approaches to this complicated issue." Id. at 4. This "complicated issue" is a manifestation of the temporal context in which this deed was executed. We therefore briefly review the two widespread and related faulty conceptual culprits referenced in Hysaw —the estate-misconception theory and the use of 1/8 as the standard royalty—that worked in tandem to lead parties to use the term "1/8" to describe something other than a literal eighth. True, it matters less why the term was used in a particular way than that it was so widely used, but understanding the reason helps eliminate any lingering doubts.

The estate-misconception theory reflects the prevalent (but, as it turns out, mistaken) belief that, in entering into an oil-and-gas lease, a lessor retained only a 1/8 interest in the minerals rather than the entire mineral estate in fee simple determinable with the possibility of reverter of the entire estate. Id. at 10 ; Concord Oil Co. v. Pennzoil Expl. & Prod. Co. , 966 S.W.2d 451, 460 (Tex. 1998). Therefore, for many years, lessors would refer to what they thought reflected their entire interest in the "mineral estate" with a simple term they understood to convey the same message: "1/8." This widespread and mistaken belief ran rampant in instruments of this time involving the reservation or conveyance of a mineral interest—so much so that courts have taken judicial notice of this widespread phenomenon. Hysaw , 483 S.W.3d at 9–10. Therefore, the very use of 1/8 in a double fraction "should be considered patent evidence that the parties were functioning under the estate misconception." Laura H. Burney, The Regrettable Rebirth of the Two-Grant Doctrine in Texas Deed Construction , 34 S. Tex. L. Rev. 73, 90 (1993). We made this point in Hysaw , citing "commentators" who have observed that "there is ‘little explanation’ for the use of double fractions to express a fixed interest absent a misunderstanding about the grantor's retained ownership interest or use of 1/8 as a proxy for the customary royalty." 483 S.W.3d at 10–11.

The second rationale—the special meaning that 1/8 acquired in the standard-royalty context—provides an additional objective indication of what parties meant by using 1/8 within a double fraction. We said in Hysaw that we had "no doubt" that "[t]he near ubiquitous nature of the 1/8 royalty—dubbed by some as ‘the legacy of the 1/8 royalty’ or ‘historical standardization’ "—is something that "influenced the language used to describe the quantum of royalty in conveyances of a certain vintage." Id. at 9–10. This prevalent belief and confusion resulted in parties mistakenly assuming the landowner's royalty would always be 1/8. Therefore, parties would use the term 1/8 as a placeholder for future royalties generally —without anyone understanding that reference to set an arithmetical value. Once again, this pervasive misunderstanding long ago led us to take judicial notice in this specific context of double-fraction royalties. Garrett , 299 S.W.2d at 907.

Working in tandem, these widely recognized principles provide objective indications about what the parties to this deed meant by deploying a double fraction. At that time, the fraction 1/8 had various meanings that linked to the landowner's conception of the entirety of the estate. We are, frankly, not aware of double fractions including 1/8 that were aimed at simple multiplication rather than referencing the mineral estate as a whole. But that does not mean it could not have happened. To account for the possibility that parties to an instrument may have intended nothing more than multiplication, the Court in Hysaw rejected a one-size-fits-all arithmetical solution in which deeds or other documents using 1/8 would inexorably be treated as referring to the entire mineral estate, as alluring as that (or any other bright-line) solution would be. Instead, we required a full contextual analysis of an instrument when construing a relevant mineral conveyance of that era, like the deed in the present case. See Hysaw , 483 S.W.3d at 10–12.

3

We now reaffirm Hysaw and clarify the import of our holding in that case. Specifically, when courts confront a double fraction involving 1/8 in an instrument, the logic of our analysis in Hysaw requires that we begin with a presumption that the mere use of such a double fraction was purposeful and that 1/8 reflects the entire mineral estate, not just 1/8 of it. Indeed, it would be odd to say "one-half of one-eighth" rather than simply "one-sixteenth" if all that was intended by "one-half of one-eighth" was 1/16. See id. (discussing the need "to accord ... significance to the use of double fractions in lieu of a single fraction"). Our analysis in Hysaw thus warrants the use of a rebuttable presumption that the term 1/8 in a double fraction in mineral instruments of this era refers to the entire mineral estate. Because there is "little explanation" for using a double fraction for any other purpose, id. at 10, this presumption reflects historical usage and common sense.

At the same time, Hysaw clarifies that the presumption is readily and genuinely rebuttable. As we made clear in that opinion, every contract, deed, or will is free to manifest a different intent and to define terms in different ways. See id. at 14 ("all the other language in the document must be considered to deduce intent"). Any instrument may be unique, and the judicial role is to ensure that individual parties receive a faithful interpretation of their legal instrument. Thus, the entire instrument should be examined to determine whether its text rebuts the presumption. We emphasize that the rebuttal must come from the document itself, given the public's right to rely on the meaning of deeds and our general insistence on predictability and stability in matters of interpretation.

Frequently, of course, examining the entire instrument will reveal provisions that confirm the presumption. Often, those additional provisions would make sense only if 1/8 is read as a term of art, as was true in Hysaw itself. No one has presented us with examples of parties to instruments of the relevant era who used a double fraction just for its arithmetical purpose, but courts should be ready to find not just confirmation but contradictions of the presumption. A rebuttal could be established by express language, distinct provisions that could not be harmonized if 1/8 is given the term-of-art usage (the mirror image of Hysaw ), or even the repeated use of fractions other than 1/8 in ways that reflect that an arithmetical expression should be given to all fractions within the instrument. In such cases, the rebuttal may be sufficiently clear that, as a matter of law, the double fraction can only be held to require simple multiplication. The key point is that there must be some textually demonstrable basis to rebut the presumption.

Nor do we foreclose the possibility that an instrument may have enough textual evidence to drain confidence in the presumption yet insufficient evidence for a court to conclude that a reasonable reader at the time would have understood the instrument to require mere multiplication. In such a case, and if our ordinary rules of construction are incapable of generating a single answer, then our case law involving inescapable ambiguity—including the authorized but reluctant recourse to extrinsic evidence—provides the next step. When that happens, a factfinder may be needed to finally resolve the text's meaning. Courts should endeavor to give meaning to the text without too hastily finding ambiguity, but there may be times in which no other choice remains.

In short, in Hysaw we rejected both available Procrustean solutions—insisting on always multiplying the double fraction or insisting on never doing so—even though the attraction of adopting either is self-evident. The approach that we describe today avoids the arbitrary, unfair, and inaccurate results that would necessarily follow from adopting either extreme. But it also avoids reinventing the wheel in every individual case. This rule is thus flexible even as it advances stability and predictability from case to case. Wasteful litigation involving double fractions seems destined to proliferate absent clear guidance for the courts and the public.

4

Given these principles, we cannot affirm the judgment below. The court of appeals concluded that it had no basis to do anything other than multiply the fractions. The estate-misconception theory played no role, that court concluded, because there were no inconsistencies to reconcile within the deed and because the property was not burdened by a lease at the time of conveyance (or at any time prior). 647 S.W.3d at 907–08.

We respectfully disagree with this analysis because we think it misapprehends how the estate-misconception theory affects the reading of instruments like the deed in this case. Specifically, the theory's relevance has never depended on the considerations that the court of appeals identified. We agree, of course, that this case somewhat differs from Hysaw and others that we have analyzed; such cases often did require a harmonization of conflicting provisions within the text. But we have never suggested a default rule that requires multiplication unless it can be proven that doing so would contravene some other provision of the text. Such a rule would ahistorically require us to read many double fractions as pointless—as though the parties just did not know how to complete a simple exercise in multiplication. As we explained in Hysaw , and as we reiterated above, the very use of the double fraction is itself the primary reason to presume purposefulness. And the temporal context of this deed indicates that those factors are clearly present here.

Accordingly, identifying a lack of inconsistent provisions that require harmonization gets the analysis backwards. The use of a double fraction in this deed, combined with the lack of anything that could rebut the presumption, is precisely why we can conclude as a matter of law that this deed did not use 1/8 in its arithmetical sense but instead reserved to the Mulkey grantors a 1/2 interest in the mineral estate.

B

Even if we were less persuaded by the double-fraction analysis, however, the Mulkey parties argue that we still would have to recognize their present-day ownership of one-half of the mineral estate. That is, even if the "one-half of one-eighth" reservation meant only a 1/16 mineral interest in 1924, they argue that the record conclusively establishes that they acquired the other 7/16 interest through the presumed-grant doctrine. The court of appeals disagreed, concluding that the presumed-grant doctrine played no role. 647 S.W.3d at 908–10.

We agree with the Mulkey parties. The presumed-grant doctrine, "also referred to as title by circumstantial evidence, has been described as a common law form of adverse possession." Fair v. Arp Club Lake, Inc. , 437 S.W.3d 619, 626 (Tex. App.—Tyler 2014, no pet.). The doctrine requires its proponent to establish three elements: (1) a long-asserted and open claim, adverse to that of the apparent owner; (2) nonclaim by the apparent owner; and (3) acquiescence by the apparent owner in the adverse claim. Magee v. Paul , 110 Tex. 470, 221 S.W. 254, 257 (1920).

The court of appeals imposed an additional fourth element: a gap in the title. 647 S.W.3d at 909. Satisfying the doctrine is properly difficult, but neither our precedent nor the doctrine's underlying purposes support mandating this additional test. Nonetheless, the extensive history of transactions and dealings detailed below, accompanied by the 1946 letter, provides enough evidence for the existence of such a gap even if it were needed.

We think that the parties’ history of repeatedly acting in reliance on each having a 1/2 mineral interest conclusively satisfies the presumed-grant doctrine's requirements. This ninety-year history includes conveyances, leases, ratifications, division orders, contracts, probate inventories, and a myriad of other recorded instruments that provided notice. There was a long and asserted open claim—for nearly a century, both parties acted in accordance with each side owning a 1/2 interest. And until this litigation began in 2013, the White parties never said anything to the contrary.

Indeed, this evidence long predates and continually postdates the 1946 letter. That fact simultaneously supplies evidence for the presumed-grant doctrine if it is actually needed and adds to our confidence that our reading of the deed is correct. For example, in 1926—just two years after the deed's execution—the White parties acknowledged in a Purchase and Escrow Agreement with an oil company that they owned only half of the minerals and that the other half belonged to the Mulkey parties. The corresponding documents provide further support by indicating that after both parties leased their respective mineral interests, each side was to receive an equal share of the royalties.

The White parties point to another transaction occurring in 1926 in which the Mulkey parties conveyed a 1/16 interest in the minerals to a third party. That 1/16 interest, however, refers to the mineral interest of an entirely different property, not any property subject to this lawsuit. We cannot see how a wholly distinct transaction sheds any light on the one before us. And our doubts about relevance are heightened because that 1926 transaction illustrates that the very same grantors could (and did) refer to a 1/16 interest directly and not via the circumlocution of a double fraction. When what they actually wanted was to convey exactly and mathematically a 1/16 interest, that is precisely what they did.

Regardless, the parties’ history marches on, particularly around the time of the 1946 letter and beyond. For example, the very week of the 1946 letter, the Mulkey parties executed an "Agreement for Division of Rentals and Royalties" with multiple recitals stating that they "own[ed] an undivided one-half interest in said minerals ...." Whether this was undertaken in conjunction with the supposed "contract" is unknown—but that it was undertaken is all that matters.

Further, again around the time of the letter, both parties leased their respective mineral interests to the same lessee, and the leases once again reflected that each party owned 1/2 of the mineral estate. The White parties filed their own recorded stipulation stating the amount of delay rentals to which they were entitled under the lease; again, the amount was consistent with the reflection that they claimed only a 1/2 interest of the mineral estate rather than the 15/16 interest they now assert. On top of this, several accompanying documents, including the division of interests and rentals, relayed the same understanding. Similar transactions occurred again in the late 1950s, including when G. R. White himself recited that he owned only an "undivided one half (1/2) interest" in "all of the oil, gas, and other minerals" on the ranch. And a series of conveyances in 1973 revealed the same thing.

These transactions continued until 2012, when Endeavor Energy drilled a well that established oil and gas production on the subject ranch land and paid both parties their respective royalty interests—1/2 to the Mulkey parties and 1/2 to the White parties. Some of the White parties then, for the first time, alleged ownership of 15/16 of the mineral estate. But the filing of this lawsuit in 2013 cannot negate nearly a century of overwhelming evidence that the White parties never previously made such a claim in all those years. In fact, all the transactions suggest that the White parties understood and intended the Mulkey parties to have a 1/2 interest. Accordingly, if the presumed-grant doctrine were in fact necessary, we would find that the Mulkey parties have conclusively established it.

III

We conclude that the Mulkey parties hold title to 1/2 of the mineral estate because the original deed so requires and because the presumed-grant doctrine would remove any remaining doubts. Because the court of appeals held otherwise, we reverse its judgment and remand to the trial court for further proceedings.