The "to-do" list of what an attorney should do if a case is reversed on appeal and sent back for retrial can be a long one. Until a recent Court of Appeal decision, further discovery would have been at, or near, the top of any post-appeal checklist.

California law provided an unfettered opportunity to conduct additional discovery after remand, with a new discovery timeline automatically recalibrating based on the date set for retrial. See Beverly Hospital v. Superior Court, 19 Cal.App.4th 1289 (1993). But the Supreme Court's grant of review in Fairmont Ins. Co. v. Superior Court, 66 Cal.App.4th 1294 (1998), may put a substantial crimp in post-appeal discovery. If the Supreme Court embraces the result in Fairmont, further discovery will be available after remand only on a showing of good cause.

The Fairmont-Beverly Hospital split of authority centers on the interpretation of the term "initial trial date" -- the anchor date on which discovery deadlines are based. See Code of Civil Procedure Section 2024(a) (discovery should be completed on or before 30th day before "date initially set for trial"); see also C.C.P. Sections 2024(e), 2030(c) (8) and 2034(a), (b) & (c) (all premising discovery deadlines on initial trial date).

Beverly Hospital liberally construed the term "initial trial date," ruling that multiple initial trial dates could exist when an action is restarted after a trial court grants a motion for new trial or declares a mistrial, or an appellate court reverses a judgment. Under the Beverly Hospital court's construction, a new initial trial date automatically recalibrates the discovery deadline.

Beverly Hospital involved a retrial after a mistrial in a medical-malpractice action. When the plaintiff served new discovery before the retrial, the defendant argued that it was untimely because the discovery deadline, calculated with reference to the original initial trial date, had long since passed.

Citing the dictionary definition of "initial" as "first," the defendants argued that an action may have only one initial trial date, no matter how many retrials eventually follow. They also argued that the trial court's discretionary power to reopen discovery for good cause offset any unfairness of this strict interpretation. The plaintiff argued for a liberal construction, pointing out that if there can be more than one trial in an action, there can also be more than one initial trial date, each date corresponding to its respective scheduled trial.

After evaluating these arguments, the court found that the phrase "initial trial date" was ambiguous: It could mean the very first trial date set in the action, or it could mean the first date set for any retrial in the action. Consistent with recognized principles of statutory construction, e.g., California Mfrs. Ass'n v. PUC, 24 Cal.3d 836 (1979), the court therefore sought to resolve this ambiguity by looking to the intent of the Legislature in enacting the statute.

But the statute's legislative history provided no definitive guidance. Apparently, the Legislature used the phrase "initial trial date" simply to clarify that a trial postponement or continuance did not automatically reopen discovery. Nothing indicated that it had ever considered retrial after a mistrial, new trial order or reversal on appeal, and no language reflected an intent to prohibit additional discovery in those contexts.

With no help coming from the legislature, the Beverly Hospital court adopted a pragmatic approach in construing the language of Section 2024. The court reasoned that permitting additional discovery was consistent with discovery's recognized purpose of facilitating trial preparation in furtherance of the interests of justice. See, e.g., Greyhound Corp. v. Superior Court, 56 Cal.2d 355 (1961).

Moreover, automatically reopening discovery would result in no unnecessary expense or delay. Rather, it would efficiently promote more focused discovery (because of what was learned before the case necessitated retrial) and possibly facilitate settlement. And automatic revival of discovery deadlines would relieve the courts and parties of the burden resulting from motions, oppositions and hearings concerning whether good cause existed to reopen discovery.

Beverly Hospital thus found that "the phrase 'initial trial date' should be interpreted to refer to the first date set for each trial in the action, not to the first trial date set in the action." Subsequent cases followed this rule, and various practice guides embraced it as well. See Province v. Center for Women's Health, 20 Cal.App.4th 1673, reversed on other grounds, Heller v. Norcal Mutual Ins. Co., 8 Cal.4th 30 (1994); 2 "Civil Discovery Practice" Sections 2.15, 2.18, 11.7 (CEB 3d ed. 1998); Robert I. Weil & Ira A. Brown, "Civil Procedure Before Trial" Discovery Sections 8:451, 8:1645 (The Rutter Group 1998).

Despite the weight of authority, the Fairmont court took a different route. Fairmont involved a plaintiff's effort to obtain additional discovery following a reversal in an insurance bad-faith action. After remand, the plaintiff served a series of discovery requests seeking production of Fairmont's file in the matter, responses to interrogatories and requests for admission.

Just like the defendants in Beverly Hospital, Fairmont objected, arguing the discovery was untimely because it was served after the discovery cut-off date -- 30 days before the original trial date. The trial court overruled Fairmont's objection and granted the plaintiff's motion to compel based on Beverly Hospital's holding that reversal of a judgment automatically restarts discovery. Fairmont sought writ relief, and the Court of Appeal reversed, rejecting Beverly Hospital's theory of multiple initial trial dates.

The Fairmont court strictly construed the phrase "initial trial date" under a plain-language analysis. Just as Beverly Hospital had done, the Fairmont court started with the dictionary and likewise found that the usual and ordinary meaning of "initially" was "first." The court did not believe, however, that there could be multiple "firsts" -- "first" meant first -- and that equated with the original trial date without the need to look further. Fairmont thus rejected Beverly Hospital's recourse to Legislative intent and characterized that opinion's construction allowing multiple initial trial dates as "strained."

Nevertheless, the Fairmont court too added its own pragmatic observations in support of its plain-language construction. Rather than viewing automatic additional discovery as a convenience to the courts and parties, the court feared that renewed unlimited discovery could result in discovery abuses. Because the parties should have completed all discovery before the first trial, there should be no need to automatically restart discovery after reversal.

The court noted that if further discovery were needed, the trial judge would have discretion to allow it by waiving the discovery cutoff on an appropriate showing of good cause. Code of Civil Procedure Section 2024(e). Requiring a showing of good cause to reopen discovery would, in turn, enhance a trial court's ability to schedule prompt and realistic dates for retrial.

With Beverly Hospital and Fairmont in direct contradiction, litigants with a scheduled retrial were unsure whether they could serve discovery based on an automatic renewal of deadlines or whether a motion seeking permission was required. The issue accordingly was perfectly positioned for Supreme Court review. California Rule of Court 29(a)(1) (securing uniformity of decision is primary reason to grant review). The Supreme Court took the case, and an opinion settling this issue is possible later this year. Meanwhile, Beverly Hospital remains the law because the high court's grant of review robs Fairmont of its citability. California Rules of Court 976(d), 977(a).

Both Beverly Hospital and Fairmont had a common goal: efficiency leading to a prompt retrial. And both had a common starting point: the recognition that clearly defined discovery time limits expedite and facilitate trial preparation and prevent delay. But the courts headed in opposite directions to reach that goal.

The Beverly Hospital court took a reasonable position: Renewed discovery would be more focused, intelligent and efficient and would not cause delay, and requiring a motion would inject delay. But the Fairmont position is reasonable too: Renewed discovery creates the potential for abuses and insisting on a showing of good cause for any further discovery limits that potential.

On balance, however, the Beverly Hospital rule seems more consistent with the purposes of pretrial discovery and the administration of justice. First, for whatever reason, all critical avenues of discovery are not always explored before trial in many cases. Renewed discovery could result in new claims or defenses or more realistic damage awards.

Second, any potential for abuse can be mitigated by the protections and time limits in the existing discovery statutes. Relying on the existing statutory scheme, the trial court can manage the burdens of discovery as it would in any other case.

Finally, the rule in Beverly Hospital reduces the possibility of inconsistency by providing a uniform standard applicable to all litigants. The routine reopening of discovery avoids the disparate treatment that could result from particularized notions of good cause.