Piney Mountain Coal Company petitions for review of an order of the Department of Labor's Benefits Review Board (BRB) affirming the decision of an administrative law judge (ALJ) to award survivor's black lung benefits to both the widow and a former spouse of James R. Mays. We affirm.

James R. Mays worked in the coal mines for forty years. He died on March 17, 1991, at the age of 67; his attending physician, Dr. E. T. Tolosa, listed carcinoma of the pancreas with metastases as the cause of death. Mays's widow, Shirley Mays, and his ex-wife, Betty Jean Mays, both filed timely applications for survivor's benefits under the Black Lung Benefits Act, 30 U.S.C. § 901 et seq. (the Act). The claims were initially denied, but the ALJ awarded benefits on August 9, 1993, after a hearing. The ALJ found that Mays had had simple pneumoconiosis arising from his four decades of coal mine employment and that pneumoconiosis had hastened his death. Pursuant to this decision the District Director issued notices directing respondent Piney Mountain Coal Company, the "responsible operator" as defined at 20 C.F.R. §§ 725.492 and 725.493, to pay full widow's benefits to both Shirley and Betty Mays. Piney Mountain then appealed the order awarding benefits, including the District Director's calculation of them, to the BRB.

After briefing and oral argument, the BRB unanimously ruled that full benefits are payable to each "widow" a miner may leave, even if there is more than one. The BRB upheld the underlying award of benefits by a split vote, ruling 2-1 that substantial evidence supported the ALJ's finding that pneumoconiosis hastened Mays's death. Mays v. Piney Mountain Coal Co., 21 BLR 1-59 (BRB 1997).

Piney Mountain has petitioned for our review.

The standard of review is a familiar one: we must affirm the decision of the ALJ if it is in accordance with law and is supported by substantial evidence. See 30 U.S.C. § 932 (incorporating 33 U.S.C. § 921(b)(3)); Doss v. Director, OWCP, 53 F.3d 654, 658 (4th Cir.1995); Wilson v. Benefits Review Bd., 748 F.2d 198, 199-200 (4th Cir.1984). "Substantial evidence is more than a mere scintilla, and must do more than create a suspicion of the existence of the fact to be established." NLRB v. Columbian Enameling & Stamping Co., 306 U.S. 292, 300 (1939). Rather, in light of the whole record, it must be of sufficient quality and quantity " 'as a reasonable mind might accept as adequate to support' " the finding under review. Richardson v. Perales, 402 U.S. 389, 401, 91 S. Ct. 1420, 28 L. Ed. 2d 842 (1971) (quoting Consolidated Edison Co. of New York v. NLRB, 305 U.S. 197, 229, 59 S. Ct. 206, 83 L. Ed. 126 (1938)). In referring to a singular "reasonable mind," the Supreme Court has directed us to uphold decisions that rest within the realm of rationality; a reviewing court has no license to " 'set aside an inference merely because it finds the opposite conclusion more reasonable or because it questions the factual basis.' " Doss, 53 F.3d at 659 (quoting Smith v. Director, OWCP, 843 F.2d 1053, 1057 (7th Cir.1988)).1

The claims before us are subject to the permanent regulations at 20 C.F.R. Part 718.2 Under those regulations claimants for survivor's benefits must show that the miner had "pneumoconiosis" as it is defined in the regulations, that the pneumoconiosis arose from coal mine employment, and that the miner's death was "due to" the disease. See 20 C.F.R. §§ 718.201 (definition of pneumoconiosis), 718.202 (existence of disease), 718.203 (relationship with coal mine employment), and 718.205 (survivor's claims). The claimant has the burden of proof by a preponderance of the evidence on all elements. 5 U.S.C. § 556(d); Director, OWCP v. Greenwich Collieries, 512 U.S. 267, 114 S. Ct. 2251, 129 L. Ed. 2d 221 (1994).

The final element of entitlement is the only one at issue in this case. According to the regulations, a death is "due to" pneumoconiosis if the disease was the primary (§ 718.205(c)(1)) or a "substantially contributing" (§ 718.205(c)(2)) cause of death, or if the irrebuttable presumption of § 718.304 is applicable.3

However, survivors are not eligible for benefits where the miner's death was caused by a traumatic injury or the principal cause of death was a medical condition not related to pneumoconiosis, unless the evidence establishes that pneumoconiosis was a substantially contributing cause of death.

20 C.F.R. § 718.205(c)(4).

Just what is a "substantially contributing cause?" The Director has adopted a temporal approach in interpreting the phrase. If pneumoconiosis actually serves to hasten death in any way, pneumoconiosis is a "substantially contributing cause." We have approved this interpretation, see Shuff v. Cedar Coal Co., 967 F.2d 977, 979 (4th Cir.1992), cert. denied, 506 U.S. 1050, 113 S. Ct. 969, 122 L. Ed. 2d 124 (1993), and we will apply it here.4

A brief synopsis of Mays's final illness will bring the legal and factual issues into focus. Mays was admitted to Tazewell Community Hospital in Tazewell, Virginia, in February 1990. He was suffering from jaundice, and a computed tomography (CT) scan of his abdomen revealed an obstructing mass at the head of the pancreas. Upon performing a cholecystojejunostomy to bypass the obstruction, Mays's doctors discovered inoperable metastatic adenocarcinoma. He was treated with radiation and chemotherapy, but he grew progressively weaker and more jaundiced as the cancer spread through his liver. He lived just thirteen months following diagnosis. The ultimate issue before the ALJ was whether "pneumoconiosis" as defined in the Act actually hastened Mays's death, notwithstanding that he would have died eventually of cancer, and the ultimate issue before us is whether the ALJ's finding that it did is in accordance with law and supported by substantial evidence.5

The primary pieces of evidence are an autopsy report, slides of lung tissue preserved from the autopsy, and the reports of consulting doctors who reviewed the slides and Mays's medical records. There is little disagreement among the doctors as to what the slides show. On the other hand, their opinions about the course of Mays's demise differ significantly.

Dr. Mario Stefanini conducted the autopsy. On gross examination, he found:

Larynx, trachea, bronchi and bronchioles contain non[-] expectorated foamy mucus mixed with debris. Pulmonary arterial tree appears grossly free. Right and left pleural cavities contain 600 and 700 ml. of turbid fluid respectively. Right and left lung[s] weigh 510 and 497 gms, respectively. Pleural surfaces show multiple blackish lines which extend within the parenchyma and numerous white subpleural nodules, 0.3 X 0.4 cms. Emphysematous bullae are noted at both apices. Serial sections show consolidation of right middle and lower lobes, with areas of increased consistency and airlessness, [and] numerous hard and blackish nodules, as large as 0.4 X 0.4 cms.

In addition, Dr. Stefanini noted "regurgitated dark debris" in Mays's esophagus and that the liver was "peppered by round, whitish nodules averaging 0.7 X 0.6 cms." The pancreas could not be evaluated because of "marked subhepatic fibrosis."

Microscopic examination of the lungs revealed numerous pathologies, many related to Mays's pancreatic cancer:

[There is] thickening of pleural surfaces; bronchi appear essentially unremarkable; there are foci of adenocarcinoma ... which are subpleural, parenchymal, perivascular and appear as tumor emboli; there is content in bronchioles of mucus secretion as well as debris[,] and surrounding confluent atelectases are numerous; ... calcified nodules are noted throughout the parenchyma; ... there are thrombi, recent, in the smallest segments of the pulmonary arterial tree and infarcts are correspondingly present; there is marked subpleural anthracosis extending into the parenchyma; there is marked peribronchiolar and perivascular anthracosis also; fibroanthracotic nodules are numerous[,] and conglomerates are occasionally present; [and] there are massive tumoral intraparenchymal metastases in areas.

Metastases were present as well in Mays's bone marrow, lymph nodes, and liver. Dr. Stefanini's "final anatomical diagnoses" included, among others, adenocarcinoma of probable pancreatic origin with metastases; failure to expectorate mucus and aspiration of debris in the airway; marantic thrombosis of the pulmonary arterial tree; chronic obstructive pulmonary disease with interstitial fibrosis and bullous emphysema; and "[c]oal worker's pneumoconiosis, complicated (conglomerates)" with fibroanthacosis of the mediastinal lymph nodes.

The mechanism of death seemed to have been related to acute respiratory failure when the patient became unable to expectorate respiratory secretions and also aspirated gastric content into the airway. The process of respiratory failure was aggravated by the development of marantic thrombosis of the pulmonary arterial tree and by the presence of pleural effusions, which seemed due to the presence of extensive pleural metastases of the primary tumor.

The autopsy had been requested to establish evidence of coal worker's pneumoconiosis. There was macronodular disease with anthracosis and the formation of a number of conglomerates.

Shirley Mays's attorney then apparently asked Dr. Stefanini to clarify what role pneumoconiosis had played in Mr. Mays's death. In a letter, dated June 25, 1991, Dr. Stefanini said:

Mr. Mays suffered from complicated coal worker's pneumoconiosis with the formation of multiple conglomerates.

It cannot be stated that coal worker's pneumoconiosis was the cause of death in this case as the patient expired from adenocarcinoma, most likely of pancreatic origin, metastatic to multiple organs and causing obstructive jaundice. However, the patient's immediate cause of death was failure to expectorate mucus with confluent atelectases of[the] lungs, complicated also by terminal thrombosis of the pulmonary arterial tree with multiple recent infarcts of [the] lungs. In this condition the presence of additional pulmonary pathology as represented by coal worker's pneumoconiosis of the complicated type could be considered a complicating factor in the demise of Mr. Mays.

(Emphasis added.)

The remainder of the record consists of critiques of Dr. Stefanini's observations and opinions. Dr. Richard Naeye reviewed the autopsy report and tissue slides at the behest of the Department of Labor. He saw "a large amount of black pigment in these lungs," including several anthracotic macronodules with a diameter of greater than seven millimeters. Focal emphysema was present, and centrilobular emphysema was "severe."6 "The most striking abnormality in the slides [was] the presence of many nests and nodules of poorly differentiated carcinoma." Dr. Naeye concluded:

A severe simple coal worker's pneumoconiosis is present. It is characterized by the presence of many anthracotic deposits, a few of which reach macronodular size. There [are] associated fibrous, birefringent crystals and focal emphysema. Assuming that the lung sections provided for my review are representative of the lungs as a whole[,] the pneumoconiosis may have been severe enough to have prevented this man from doing hard physical work in the coal mining industry. This would be difficult to substantiate with the results of pulmonary function studies because centrilobular emphysema is severe.

Dr. Echols Hansbarger reviewed the medical records and slides at the request of the employer. His description of the slides was much like Dr. Naeye's:

Lung: Moderate emphysema is noted with enlarged air spaces and chronic inflammation of bronchi. Some collapse of lung parenchyma is seen with atelectasis. Several areas are noted in the lung that show diffuse infiltration by a poorly differentiated malignant neoplasm composed of apparent glandular cells with focal necrosis. Scattered throughout the lung parenchyma in addition are numerous deposits of anthracotic pigment with reactive fibrosis. Many small coal maculae are noted with a range in size from 0.1 cm to 0.3 cm. Some larger areas are noted with a dense fibrotic center. These measure up to 1 cm in greatest size. Some pleural fibrosis is noted with pigment deposition.

Dr. Hansbarger concluded that Mays "died as a result of Adenocarcinoma of the Pancreas with widespread systemic metastatic disease." Like Dr. Naeye, he noted "rather severe" simple coal worker's pneumoconiosis "of the Nodular Fibrosis type." Finally, he said, "I do not believe the [pneumoconiosis] ... contributed to his demise in a major way."

In a subsequent letter to the Department of Labor, Dr. Hansbarger explained his opinion in greater detail. He first defined his terms. He deemed "anthracosis" to be carbon retention in the lungs without tissue reaction, and he said that the condition is "very common ... in urban society." He divided simple coal workers' pneumoconiosis into the "Dust Reticulation Type" (coal maculae of less than seven millimeters in diameter) and the "Nodular Fibrosis Type" (seven millimeters to two centimeters in diameter). Nodules of over two centimeters were necessary "[b]y definition" to justify a diagnosis of complicated pneumoconiosis or "progressive massive fibrosis." In support of his definitions, he cited a "classic monograph" on the disease: Jerome Kleinerman, et al., Pathology Standards for Coal Workers' Pneumoconiosis, 103 Archives of Pathology & Laboratory Medicine 374 (1979). According to this monograph, excerpts of which were attached to Dr. Hansbarger's report, the two-centimeter minimum diameter for complicated pneumoconiosis is an "arbitrary choice" that "does not imply that smaller lesions do not occur," but rather "permits better correlation with clinical and roentgenographic measurements." Id. at 379.

Applying this definition, Dr. Hansbarger reiterated that Mays had suffered from severe, but "simple," pneumoconiosis. He then spoke directly to the ultimate issue:

I definitely feel that there was no contribution to this individual[']s death because of the Simple Coal Workers' Pneumoconiosis. This is because of the extensive wide spread metastatic carcinoma which caused his death. I do not think the presence of this Coal Workers' Pneumoconiosis hasten[ed] this demise in any way.

Dr. Emery Robinette examined the medical record at the request of Shirley Mays. He, too, saw Mays's pneumoconiosis as an intermediate form, more severe than the clinical label "simple" might suggest:

[Mays's] coal workers' pneumoconiosis is morphologically classifiable as simple coal workers' pneumoconiosis, but it must be recognized that [he] had palpable black nodules that were smaller than 1 cm. in diameter. These nodules are described in many miners at autopsy[,] and it is important to realize that on microscopic examination these nodules blend in perceptively [sic] with the coal macule on one hand and with the areas of progressive massive fibrosis on the other. It is likely that these nodules represent steps in the spectrum of changes rather than a pathologically distinct lesion. These nodules are probably the precursor of areas of progressive massive fibrosis. Nodular lesions of greater than 1 cm. in diameter are arbitrarily termed progressive massive fibrosis ... although there is no significant pathological difference between nodules less than 1 cm. in size versus 1 cm. in size....

Dr. Robinette then noted that Mays had suffered from mild resting hypoxemia, and he opined that Mays's shortness of breath was attributable to pneumoconiosis. He offered no opinion on whether the disease hastened death, though he did point out that there is no relationship between pneumoconiosis and pancreatic cancer.

Dr. Jerome Kleinerman reviewed the record at the instance of Piney Mountain. On the tissue slides, he noted a "moderate number"of coal macules in "virtually all sections" and several small (three to eight millimeters in diameter) nodules of simple nodular silicosis. He concluded that Mays's death was caused by adenocarcinoma of the pancreas with liver and lung metastases, which condition was "in no way" related to coal dust exposure.7

After summarizing the above evidence, the ALJ first concluded that the irrebuttable presumption of death due to complicated pneumoconiosis was not applicable. He found that pancreatic cancer was the primary cause of death. He then emphasized, however, that Dr. Stefanini's "gross autopsy descriptions are unchallenged." These descriptions show that Mays's lungs and air passages were filled with mucus and debris. Thus, the mechanism of death was Mays's inability to expectorate mucus. The ALJ, of course, relied on Dr. Stefanini's June 25, 1991, letter as an opinion that pneumoconiosis was a complicating factor in Mays's death. Finally, the ALJ pointed to Dr. Naeye's description of the severity of Mays's pneumoconiosis, combined with Dr. Hansbarger's concession that pneumoconiosis of sufficient severity could interfere with expectoration, as further inferential support for Dr. Stefanini's view. The ALJ therefore held that pneumoconiosis had hastened death.

Piney Mountain first argues that the ALJ could not and should not have credited Dr. Stefanini's opinion regarding causation after first rejecting his opinion that Mays suffered from "complicated" pneumoconiosis. In support of this argument Piney Mountain cites our decision in Grigg v. Director, OWCP, 28 F.3d 416, 419 (4th Cir.1994) (holding that the opinion regarding disability causation of a doctor who erroneously fails to diagnose pneumoconiosis is not worthy of enough weight to rebut the interim presumption pursuant to 20 C.F.R. § 727.203(b)(3)); see also Toler v. Eastern Associated Coal Co., 43 F.3d 109, 115-116 (4th Cir.1995) (adapting and applying Grigg in a case arising under the permanent regulations).

The Grigg rule is a perfectly sound one, but our later cases have explained that it must be applied with an ear sensitive to conflicting meanings ascribed to the same words by lawyers and doctors, as well as to idiosyncratic differences in phraseology among doctors themselves. Between doctors and lawyers, the most common of these conflicts is the stark disparity between the clinical definition of "pneumoconiosis" and the broader legal meaning of the same term. See, e.g., Dehue Coal Co. v. Ballard, 65 F.3d 1189 (4th Cir.1995); Hobbs v. Clinchfield Coal Co., 45 F.3d 819 (4th Cir.1995). As Dehue Coal noted, a doctor who diagnoses a chronic respiratory disease aggravated by coal dust exposure, but who denies the presence of "pneumoconiosis," has perhaps made no error at all. 65 F.3d at 1193.

Hobbs is a good example. In that case the ALJ found that the miner suffered from "pneumoconiosis" based on chronic bronchitis attributable to coal dust exposure. Several doctors who diagnosed this condition and conceded its occupational genesis also opined that the miner was not suffering from "pneumoconiosis" and that the bronchitis did not contribute to any disability. 45 F.3d at 822. The ALJ credited these doctors' opinions in denying benefits, and we affirmed, holding that the ALJ had not run afoul of Grigg. Id. at 821-822. We affirmed because the ALJ had found the same fact that the doctors had (the miner had a chronic pulmonary disease attributable to inhalation of coal dust), although the ALJ would label that fact "pneumoconiosis" and the doctors would not.

Thus, in applying the Grigg rule, our focus should be on the descriptive facts and opinions of a doctor and not upon whether his use of some medical term of art jibes with the ALJ's use of some legal term of art. Here, we are not only confronted with a clinical term, "complicated pneumoconiosis," but also a term that is not given a universal definition among the clinical opinions of record.8

Although Dr. Stefanini called Mays's pneumoconiosis "complicated," nothing in his description of Mays's lungs could lead anyone to believe that he saw the "massive lesions" required for application of 20 C.F.R. § 718.304, and the ALJ held, in effect, that those lesions were not present. Thus, there is no inconsistency between the factual findings of Dr. Stefanini and the ALJ.

On this subject, we must note that the largest nodules Dr. Stefanini described were only four millimeters in diameter, quite smaller than the seven-millimeter and one-centimeter nodules reported by Drs. Naeye and Hansbarger. If anything, then, Dr. Stefanini may have underestimated the severity of Mays's pneumoconiosis, whether it be labeled "simple" or "complicated."9 In short, the ALJ was not required to reject Dr. Stefanini's opinion based on Grigg and its subsequent refinements.

Piney Mountain next assails the substantiality of the evidence, and this argument presents us with a very close question. The company asserts that Dr. Stefanini's opinion is both uncorroborated and too equivocal to carry the claimants' burden of proof. The "substantial evidence" standard is tolerant of a wide range of findings on a given record, but the ALJ's decision here tests the limits of that tolerance.10

We are not particularly troubled by the level of corroboration. On a critical point--the gross appearance of Mays's lungs at autopsy--there was not and could not have been either corroboration or contradiction. Dr. Hansbarger conceded at deposition that he accepted Dr. Stefanini's observations on gross examination, specifically, that Mays's lungs and airways were filled with fluid and debris. Consequently, Dr. Hansbarger said that he "had no reason to doubt" Dr. Stefanini's opinion that the mechanism of May's death was respiratory failure occasioned by an inability to expectorate mucus. Finally, though Dr. Hansbarger testified that simple coal workers' pneumoconiosis does not itself produce mucus, he acknowledged (hypothetically) that if the disease is severe enough, it can hamper the expectoration of mucus produced in some other fashion.

There is therefore substantial evidence that Mays's respiratory tract failed and that he essentially drowned in his own mucus.11 Moreover, all of the doctors agree that Mays suffered from several pulmonary diseases, including severe simple coal workers' pneumoconiosis. The ALJ could reasonably infer that one or some combination of those diseases led to the inability to expectorate. The closer issue is whether substantial evidence supports a finding that pneumoconiosis was one of them.

Dr. Stefanini's actual words (in his letter of June 25, 1991) were that pneumoconiosis "could be considered a complicating factor" in Mays's death. His use of the passive voice saps the vigor of his opinion, though not the substance, inasmuch as it is clear that he is the one doing the "considering." More troublesome is his use of the conditional "could." Piney Mountain argues that we must reject his opinion as too equivocal as a matter of law to satisfy the burden of proof. In the end, we must disagree with the company.

Claimants need not prove entitlement beyond a doubt, but rather by a simple preponderance of the evidence. Assessment of the complexities of human health and disease defies death-and-taxes confidence, and we have noted that the "state of medical knowledge concerning the exact consequences of prolonged exposure to coal dust" is "uncertain." Bethlehem Mines Corp. v. Massey, 736 F.2d 120, 124 (4th Cir.1984).

Of course, uncertainty is not proof, and claimants must prove entitlement. Nevertheless, a reasoned medical opinion is not rendered a nullity because it acknowledges the limits of reasoned medical opinions. Many wise speakers choose their words carefully and conservatively, never overstating as certain an opinion that admits of any doubt, and some timid ones unnecessarily couch a sound message in noncommittal language. Still others "believe passionately in the palpably not true," and forgo no opportunity to share these beliefs.12 In sum, the reliability of a given opinion is not necessarily revealed by the forcefulness of the speaker's language.

Again, Dr. Stefanini said that pneumoconiosis "could be considered a complicating factor in the demise of Mr. Mays." It is reasonable to read Dr. Stefanini's words as simply acknowledging the uncertainty inherent in medical opinions, while nevertheless offering a positive opinion about Mays's cause of death. In the full context of his report and of the gross observations he alone made, the reasonableness of such a reading is even clearer. Dr. Stefanini was plainly impressed by both the severity of Mays's pneumoconiosis and by the respiratory failure that ushered Mays through death's door. These key underlying impressions are corroborated to the extent they can be, and to the extent they cannot be, they are uncontradicted. A reasonable reader would take them into account in assessing the intended meaning and force of the word "could."13

Moreover, if we are going to focus on single words, there is a second word of some relevance. In the very same letter, see supra, at 8, Dr. Stefanini began with an opinion that pneumoconiosis was not "the cause of death." His next word was "[h]owever," and he then discussed Mays's respiratory failure. "However" implies a contradiction, exception, or qualification to what has been said before. Pneumoconiosis was the subject of what had come before, so the entire succeeding discussion was moot unless that disease contributed to the respiratory failure discussed.

Finally, all four judges in the administrative process--even the BRB dissenter--read Dr. Stefanini's opinion as an affirmative assertion that pneumoconiosis contributed to Mays's death.14

Of course, both the meaning of an ambiguous word or phrase and the weight to give the testimony of an uncertain witness are questions for the trier of fact. See, e.g., Pegasus Helicopters v. United Technologies Corp., 35 F.3d 507, 511 (10th Cir.1994) (meaning of ambiguous contract term was question of fact); 29A Am.Jur.2d Evidence § 1442 (1994 & Supp.1998) (To be credited by a trier of fact, a "witness is not required to speak with such confidence as to show that there are no doubts in the witness's mind."). Thus, we must review the ALJ's reading of Dr. Stefanini's opinion through the prism of the "substantial evidence" standard. This case is a close call, as we said in the beginning. But to overturn the ALJ, we would have to rule as a matter of law that no "reasonable mind" could have interpreted and credited the doctor's opinion as the ALJ did. That we cannot do. The award of benefits is supported by substantial evidence and must be affirmed.

Finally, Piney Mountain argues that if Mays's death was due to pneumoconiosis, full survivors' benefits should not be payable to both Shirley Mays and Betty Jean Mays, even though both met the Act's definition of "widow."15 The Director has participated in this case in order to urge us to uphold his interpretation of the statute and regulations as providing for full benefits to each widow. We review the BRB's application of the law de novo. See Scott v. Mason Coal Co., 60 F.3d 1138, 1140 (4th Cir.1995). On the other hand, the BRB here followed the Director's interpretation of the law, and we must defer to that interpretation if it is reasonable. See Pauley v. BethEnergy Mines, Inc., 501 U.S. 680, 696, 111 S. Ct. 2524, 115 L. Ed. 2d 604 (1991); Lisa Lee Mines v. Director, OWCP, 86 F.3d 1358, 1362 (4th Cir.1996) (en banc). We believe that it is.

If Mays were alive and collecting disability benefits in his own right, he would receive only a 50 percent augmentation of the basic benefit for one additional dependent and 75 percent for two. 30 U.S.C. § 922(a)(4); 20 C.F.R. § 725.520(c)(3). Piney Mountain asserts that the total benefits that may be paid to his survivors is limited in the same way: inasmuch as there are two eligible beneficiaries, its total liability should equal one basic benefit plus 50 percent. In support of its view Piney Mountain cites 20 C.F.R. §§ 725.533(a)(4) and 725.537.16 The Director counters that these regulations merely describe the manner in which benefit reductions are implemented, but they do not establish when reductions are warranted. A closer look at the statutory scheme causes us to agree with the Director.

The Act provides that "benefits shall be paid to[the miner's] widow (if any) at the rate the deceased miner would receive benefits if he were totally disabled." 30 U.S.C. § 922(a)(2). The term "widow" includes a "surviving divorced wife" as that term is defined by the Social Security Act, if the "surviving divorced wife" was dependent on the miner as of the month before he died. 30 U.S.C. § 902(e) (incorporating definition at 42 U.S.C. § 416(d)(2)).17

Although this definition would inevitably (if infrequently) lead to situations like this one, Congress did not prescribe an alternative method to compute multiple widows' benefits. On the other hand, it did expressly provide for sharing of an augmented basic benefit when the miner was survived only by multiple children or other dependent relatives. See 30 U.S.C. § 922(a)(3), (5). We ordinarily presume that such "disparate inclusion or exclusion" in a statute is intentional and purposeful. Russello v. United States, 464 U.S. 16, 23, 104 S. Ct. 296, 78 L. Ed. 2d 17 (1983).

The regulations repeat the distinction. A "miner" or "surviving spouse" is entitled to the full basic benefit for any "month for which he or she has no dependents who qualify under this part." 20 C.F.R. § 725.520(b) (emphasis added). On the other hand, other types of qualifying survivor are entitled to the full benefit only where the survivor "is the only beneficiary entitled to benefits." Id. (emphasis added).

Thus, the faulty premise of Piney Mountain's argument is that, as in a living miner's claim, there is a single award of benefits that should be augmented on account of the number of dependents the claimant has. In fact, a surviving widow is a beneficiary in her own right.

To adopt Piney Mountain's position, we would be forced to deem just one of the widows a primary beneficiary and the other as a dependent augmentee. But neither the statute nor the regulations tells us which "widow" is primary and which is augmentee, let alone how or whether to apportion the total benefit. "Augmentees" are "dependents" of the miner or surviving spouse, and the benefits of the "miner or surviving spouse" are augmented. 20 C.F.R. § 725.520(c)(1), (2). The regulation thus contemplates that the claimant will support his or her dependents from the augmented benefit, just as he or she should whatever the source of income. This scheme, we are certain, is not well fitted to the living arrangements commonly found among multiple ex-wives. Should the company pay Betty Mays, and expect her to support Shirley, or vice versa? Recognizing this regulatory gap, Piney Mountain suggests that the widows simply split the money. While this suggestion may be equitable, it is also manufactured out of whole cloth.

Moreover, we do know that Congress intended the definition of "widow" to be the same under the Black Lung Benefits and Social Security Acts,18 and the latter statute expressly provides for full benefits to each surviving spouse or surviving divorced spouse. 42 U.S.C. § 403(a)(3)(C). Piney Mountain acknowledges that Congress sought congruence between the statutes, but argues that the congruence ends at the bare definition of the status of "widow." The Director discerns a somewhat broader congressional intent to eliminate obscure bureaucratic distinctions between programs with similar operations and goals, and that assessment is surely not unreasonable.

Nevertheless, for a time the Director did have an internal operating procedure interpreting the Act in just the manner Piney Mountain urges.19 In 1992 this internal operating procedure was revised and the revision published.20 Then, in 1997, as a part of a comprehensive revision of the black lung regulations, the Director proposed a new paragraph to formally promulgate his current view. In proposing this regulation, the Director explained its purpose and the reasons his interpretation of the Act had changed:

Proposed [20 C.F.R. § 725.121(b) ] reflects the Department's position that the [Act] and pertinent legislative history require the payment of full monthly benefits to each surviving spouse and surviving divorced spouse who satisfies the entitlement criteria, regardless of the existence of any other spouse who also qualifies for benefits.

Prior to 1992, the Department's policy regarding the allocation of benefits between (or among) multiple surviving spouses of the same miner ... limited each spouse to [an equal share of the maximum benefit due a single claimant with the given number of augmentees]. In 1992, the Department ... concluded that each surviving spouse who meets the criteria for eligibility is entitled to the payment of the full benefits due a surviving spouse. The change of position was the result of further reflection in pertinent provisions of the [Act] and their legislative history.

62 Fed.Reg. 3338, 3350 (Jan. 22, 1997). After reviewing the legislative history, the Director concluded:

A miner, as the primary beneficiary on a claim, is clearly entitled to a full basic benefit. Upon the miner's death, the "widow," as the primary beneficiary, must be compensated in like fashion. Section 902(e) defines the term "widow" to include both a surviving spouse and a surviving divorced spouse. Nothing in § 922 provides for an alternative payment amount if a miner is survived by two widows. Consequently, the plain language of the statutory payment provisions mandates that both spouses should receive a full (100 percent) basic benefit amount. To utilize any other methodology would require payment to each "widow" at less than the statutorily prescribed "rate the deceased miner would receive if he were totally disabled."

Id. at 3350-3351 (citations omitted).

Though arbitrary changes in an agency's position may lessen the degree of deference owed that position, see Smiley v. Citibank (South Dakota), N.A., 517 U.S. 735, 742, 116 S. Ct. 1730, 135 L. Ed. 2d 25 (1996), the Director took commendable care to explain this change cogently. An agency willing to study a statute and legislative history and, where appropriate, to amend its interpretations is more deserving of the respect and deference of the courts than an agency that foolishly stays on the wrong path merely because it is well worn.

Because the Director has reasonably interpreted the Act and regulations to provide for the payment of full survivor's benefits to each "widow," we must defer to that interpretation and affirm the awards.

AFFIRMED.

NIEMEYER, Circuit Judge, concurring in part and dissenting in part:

I concur in the majority's conclusion that the Black Lung Benefits Act ("Act"), namely 30 U.S.C. § 922(a), provides for full benefits to a surviving spouse and a surviving ex-spouse. I disagree, however, with the majority's conclusion that the record contains evidence establishing that pneumoconiosis substantially contributed to James Mays' death. Accordingly, I respectfully dissent from that portion of the opinion for the court.

The majority draws upon our decision in Shuff v. Cedar Coal Co.,967 F.2d 977, 979 (4th Cir.1992), to discern the standard of liability applicable in this case but then seems to discard the requirements of the same case's holding. I believe that our approach in Shuff must guide, if not control, the resolution of this appeal.

In Shuff, much like the present case, an autopsy determined that pancreatic cancer ultimately caused a miner's death even though the miner had contracted pneumonia just before dying. The ALJ, therefore, examined the evidence to determine "whether pneumoconiosis was a substantially contributing cause of death ... insofar as it made the miner more susceptible to pneumonia." Shuff, 967 F.2d at 979. In denying an award of benefits, "[t]he ALJ concluded that while 'it appears that pneumoconiosis may have hastened ... death,' the fact that Mr. Shuff's death was 'imminent' from the pancreatic cancer, and that 'the cancer itself made Mr. Shuff more susceptible to pneumonia,' meant that pneumoconiosis was not a substantially contributing cause of death." Id. at 979 (emphasis added). We remanded the case for a "definitive finding as to whether the pneumoconiosis actually hastened" the miner's death, citing the conflicting medical evidence and the ALJ's equivocal, inconclusive statement. Id. at 980. We should do the same here.

The Black Lung Benefits Act's plain import requires proof of causation--i.e. proof "that establishes that pneumoconiosis was a substantially contributing cause of death." 20 C.F.R. § 718.205(c)(4) (emphasis added). No matter how the majority elects to spin it, Dr. Mario Stefanini's opinion--the only opinion on which the ALJ in this case could rely--can only be used for what it says: "coal worker's pneumoconiosis of the complicated type could be considered a complicating factor in the demise of Mr. Mays." (Emphasis added). That statement is much too indefinite and equivocal to pass muster under 20 C.F.R. § 718.205(c)(4). Moreover, the opinion of Dr. Echols Hansbarger does nothing to buttress the ALJ's (and majority's) conclusion regarding the cause of Mays' death. Dr. Hansbarger plainly stated, both in his written report and deposition, that pneumoconiosis did not hasten Mays' death. That Dr. Hansbarger "hypothetically" acknowledged that pneumonoconiosis could hamper the expectoration of mucus is irrelevant. See ante at 15. The Act demands a determination of causation. Hypothetical inquiries cannot serve as the basis for an award of benefits.

I am further troubled by both the ALJ's and majority's failure to mention, yet alone consider, the conclusion of Mays' treating physician, Dr. E.T. Toloso, that pancreatic cancer caused Mays' death. Certificates of death in Virginia provide space for a treating physician to designate "[o]ther significant conditions contributing to death but not resulting in the underlying cause." Dr. Toloso, significantly, left that portion of Mays' certificate of death blank. Dr. Tolosa's status as a treating physician "entitles his opinion to great, though not necessarily dispositive, weight." See Grigg v. Director, Office of Workers' Compensation Programs, 28 F.3d 416, 420 (4th Cir.1994); see also Richardson v. Director, Office of Workers' Compensation Programs, 94 F.3d 164, 168 (4th Cir.1996) (holding that remand was appropriate given ALJ's failure to weigh treating physician's opinion). In my view, the ALJ's failure to consider Dr. Toloso's opinion alone, was reversible error.

In summary, the record, as it currently exists, does not support the ALJ's conclusion that pneumoconiosis hastened and thereby substantially contributed to Mays' death. Indeed, the evidence could lead a rational fact finder to just the opposite conclusion. The prudent course of action, at the least, would be to follow our holding in Shuff and remand this case for further fact finding on causation. The record is simply inconclusive as to the role, if any, that pneumoconiosis played in causing Mays' death, and until such a determination is made, the ALJ was without authority to award benefits under the Act.