concurring.
I concur in the judgment, but I disclaim the reasoning by which the majority come to it.
The majority purport to employ a substantial-evidence analysis to uphold certain declarations in the agency order, each of which is entitled “finding of fact.” I refer to those “findings of fact” numbered six, seven, and eight, discussed in the majority opinion. Notwithstanding the name supplied by the agency, these patently are not “findings of fact” in the sense in which that term of art is used in administrative law. That is to say, the text of each shows them to be something other than a declaration of what the reality is concerning an evidentiary issue which is material to the case but without legal effect in and of itself. See generally Powers, Judicial Review of the Findings of Fact Made by Texas Administrative Agencies in Contested Cases, 16 Tex.Tech L.Rev. 475, 477-478 (1985). Applying to them the substantial-evidence test, on judicial review, is akin to checking the brakes of an automobile by reading the gasoline gauge. They require a different method of analysis, based on what they really are.
“FINDING OF FACT NO. 8”
The declaration labeled “Finding of Fact No. 8” is actually a conclusion of law which may embody as well an underlying agency policy decision. The “finding” declares that the dissolved-oxygen aspect of the receiving waters, under some conditions and at a particular point, “is already 50 percent lower than the standard” established by the agency; and this “militates against issuance of a new discharge permit that would result in more oxygen demanding material” being diverted into Dickinson Bayou. As mentioned in the majority opinion, the appellant’s concede the substandard state of the receiving waters and argue that their proposed effluent simply will not make much difference.
The word “militate” means in ordinary usage that the undisputed, existing, substandard state of the receiving waters is a factor to which the agency assigned weight or effect in reaching some determination. Thus, “finding of fact” number eight rejects on its face any idea that it might itself be a determination of any kind, whether of law or fact; and it is rather fictitious to purport to look for evidence which reason*808ably supports a finding the agency never made at all.
The actual dispute may only be whether the agency could validly make the substandard state of the receiving waters a legally relevant factor in deciding some issue pertinent to the agency’s administration of the permit procedure authorized in Tex.Water Code Ann. § 26.027 (1988). This plainly is not a question which a court may review under the substantial-evidence test of Texas Administrative Procedure and Texas Register Act (APTRA), Tex.Rev.Civ. Stat.Ann. art. 6252-13a, § 19(e)(5) (Supp. 1989). It is, instead, a question of statutory construction to be determined by an interpretation of the powers, functions, duties, and discretion assigned the agency in the Texas Water Code, particularly in Chapter 26. In other words, it is a question of law, and perhaps a complicated one at that, reviewable by a court under AP-TRA § 19(e)(1), (2), (4), or (6). The parties have not briefed the issue on that basis, having instead confined their argument to appellants’ “no evidence” theory. Accordingly, I concur that the point of error should be overruled.
“FINDINGS OF FACT” SIX AND SEVEN
The declarations labeled “Finding of Fact No. 6” and “Finding of Fact No. 7” are actually reasons given by the agency for declining to credit certain evidence adduced by the appellants in attempting to carry their burden of proof regarding specific norms and standards necessary to obtain a permit: item six states that appellants failed to establish that water quality would not be impaired in the receiving waters because their “Streeter-Phelps” analysis “failed to account for” tidal effects; item seven states that appellants failed to establish their effluent discharge would meet certain criteria because the opinion testimony given by their expert witness “was speculative and contradictory” concerning the factors of taste and odor.
The appellants contend their evidence under number seven required belief because it was uncontradicted, and I will assume that the evidence under number seven was the same. This poses the best case possible for the appellants, and requires that we consider these issues: when may an agency, as fact finder, decline to infer the facts suggested by uncontradicted evidence on the ground that the agency chooses not to believe it for some reason; and how should a court review a claim that the agency was legally bound to believe the evidence and infer the facts which the evidence implied?
“It is a generally sound rule that questions of credibility are not reached by the substantial evidence test.” Jaffe, Judicial Control of Administrative Action, at 606 (1965) [hereinafter Judicial Control]. This is implied in APTRA § 19(e)(5), which limits the scope of a reviewing court’s substantial-evidence inquiry to “the reliable and probative evidence in the record as a whole.” (emphasis added). When an agency explicitly declares, as it did here, that it could not make the findings of fact necessary to establish a required norm or standard, because it did not credit the proponent’s evidence, the court is obviously not faced with the substantial-evidence inquiry: could the fact finder reasonably infer its findings from the reliable and probative evidence in the record as a whole? Such an inquiry is simply not possible in a case like the present where the proponent contends certain findings should have been made, because his evidence was undisputed, but were not. Moreover, the reviewing court may not itself review the record as a whole and determine whether the proponent’s evidence was “reliable and probative,” for this would amount to substituting the court’s judgment for that of the agency on the weight of the evidence — a course which APTRA § 19(e) expressly forbids the reviewing court to follow. “A court addressing itself to credibility may end by ‘weighing the evidence.’ ” Judicial Control, supra, at 606.
Reviewing courts often state that the agency is the sole judge of the credibility of witnesses, which is true as a general statement; but reviewing courts do in some instances set aside agency orders which rest on an agency decision that a *809party’s undisputed evidence is not worthy of belief. Judicial Control, supra, at 606-10. The difficult issue raised in such cases concerns the exact nature of the scope of judicial review. Various epithets and metaphors abound, ranging from an agency’s “irrational” disregard of uncontradicted evidence; to its permissible disregard of un-contradicted evidence that carried “its own death wound”; and finally to statements that come perilously close to sanctioning a reviewing court’s substitution of its own judgment, for that of the agency, on the weight of the evidence. The issue is one of the more unsettled and difficult in administrative law, but in any event the scope of review is that of APTRA § 19(e)(6): Was the agency action “arbitrary or capricious,” an “abuse of discretion,” or a “clearly unwarranted exercise of discretion.” It manifestly is not a question which comes within the substantial-evidence test of APTRA § 19(e)(5), for a wow-finding simply cannot be subjected to a substantial-evidence analysis.
In the present case, the agency gave specific reasons for refusing to credit the appellants' evidence. I would hold the reasons adequate on their face and sufficiently explained. 5 Davis, Administrative Law Treatise § 29.26 at 456 (2d ed. 1984).
I believe finally that the agency would have denied the permit on the strength of its conclusion of law number four alone.
Accordingly, I would affirm the judgment below.