I concur in the result.
I agree that we must discharge the order to show cause. I do so because I believe that petitioner has not alleged specific facts that would entitle him to relief on habeas corpus.1
*816I would rest, however, solely on the merits, and would not proceed to any of the so-called “procedural bars,” including that of untimeliness.2
Like the majority, I am of the view that the writ of habeas corpus “ ‘was not created for the purpose of defeating or embarrassing justice, but to promote it[.]’ ” (Maj. opn., ante, at p. 777, quoting In re Alpine (1928) 203 Cal. 731, 744 [265 P. 947, 58 A.L.R. 1500].) That is especially so when, as here, the judgment that the petitioner challenges includes a sentence of death.
But, unlike the majority, I would follow the only path that would guarantee the attainment of justice in such a situation, namely the “examin[ation]” of “each” habeas corpus “petition on its own facts” in order to determine whether the petitioner has alleged specific facts that would entitle him to relief. (In re Clark (1993) 5 Cal.4th 750, 802-803 [21 Cal.Rptr.2d 509, 855 P.2d 729] (cone, and dis. opn. of Mosk, J.).)
To be sure, such “scrutiny” of the merits “requires the expense of considerable judicial resources . . . .” (In re Clark, supra, 5 Cal.4th at p. 803 (cone, and dis. opn. of Mosk, J.).) Consider the present petition, together with its “supplement”: It contains 40 claims, with numerous subclaims, in hundreds of pages; it is supported by 213 exhibits, in several parts, filling pages numbering in the thousands.
Scrutiny of the merits, however, requires much less than does the effort to invoke each and every procedural bar. Or even the effort to invoke any one such bar. For proof, the reader need only peruse the many pages that the majority devote to the single procedural bar of untimeliness as to a single, *817patently unmeritorious, claim among 39 others. (See cone, and dis. opn. of Kennard, J., post, at p. 819, fn. 2.)3
Because principle and pragmatism show themselves in rare conjunction in this proceeding, as in all others in which the petitioner challenges a judgment including a sentence of death (see generally, In re Gallego (1998) 18 Cal.4th 825, 842-853 [77 Cal.Rptr.2d 132, 959 P.2d 290] (cone, and dis. opn. of Brown, J.)), I would rest solely on the merits and discharge the order to show cause accordingly.
A habeas corpus petitioner must carry burdens of both pleading and proof. (See, e.g., In re Sassounian (1995) 9 Cal.4th 535, 546-547 [37 Cal.Rptr.2d 446, 887 P.2d 527].) Initially, he *816must carry the burden of alleging specific facts. (Id. at p. 546.) He must do that in his petition and, subsequently, in his traverse to any return by the respondent. (E.g., People v. Duvall (1995) 9 Cal.4th 464, 474-478 [37 Cal.Rptr.2d 259, 886 P.2d 1252].) If he succeeds, he must then carry the burden of proving such facts—to be precise, such facts as are material and subject to dispute (see id. at p. 478)—by a preponderance of the evidence. (In re Sassounian, supra, 9 Cal.4th at pp. 546-547.) He must do that at an evidentiary hearing before a referee (People v. Duvall, supra, 9 Cal.4th at p. 478), which is conducted pursuant to the rules of evidence (In re Fields (1990) 51 Cal.3d 1063, 1070 [275 Cal.Rptr. 384, 800 P.2d 862]).
The majority make a novel departure from the law respecting the burdens of pleading and proof that a habeas corpus petitioner must carry (see, ante, at p. 815, fn. 1), at least for facts relevant to the procedural bar of untimeliness and perhaps for those relevant to others as well. They require him to carry his burden of proving such facts as he attempts to carry his burden of pleading them, and to do so, outside of any evidentiary hearing, in his petition and, if applicable, his traverse. They are at once too lenient, for they allow him to “prove” facts merely by alleging them, and too harsh, for they compel him to make the attempt in ignorance of the evidentiary rules that we may or may not choose to follow in a newly fashioned procedure that may fairly be characterized as a “nonevidentiary nonhearing.”
The procedural bar of untimeliness is “indeterminate at [its] very core.” (In re Clark, supra, 5 Cal.4th at p. 802 (cone, and dis. opn. of Mosk, J.).) Its operative phrases are “substantial delay” and “good cause.” (See, e.g., maj. opn., ante, at p. 779.) Of course, its application to any given claim may yield varying results, as reasonable persons differ as to whether the claim in question has been presented without “substantial delay” and, if not, whether “good cause” exists for any such delay. But, worse yet, its very meaning is vague, as such persons attempt to discern the sense of “substantial delay” and “good cause.” And its meaning will become vaguer still, if they are called on to consider whether a claim that is timely should somehow be deemed untimely because of a “societal interest[] in finality.” (Id. at p. 794.) The majority assert that the procedural bar of untimeliness has caused petitioners to file their petitions more expeditiously. But they cannot deny that it has caused us to dispose of such petitions at an altogether different pace.