Opinion

In this action, the People of the State of California, acting through the Attorney General, succeeded in obtaining an injunction requiring the Tehama County Board of Supervisors to apply the provisions of the Subdivision Map Act (Gov. Code, § 66410 et seq.) to a lot line adjustment on property owned by defendant KAKE, LLC (KAKE). The trial court then awarded the Attorney General $173,450 in attorney fees against the county defendants under section 1021.5 of the Code of Civil Procedure, which is the codification of the private attorney general doctrine of attorney fee recovery.

On appeal from the judgment, KAKE and the county defendants contend the lot line adjustment was exempt from the Subdivision Map Act because it did not create a greater number of parcels than previously existed, and the trial court erred in concluding otherwise.

On appeal from the award of attorney fees, the county defendants and the People both contend the trial court erred in basing its award under Code of Civil Procedure section 1021.5 on the conduct of the county defendants’ former attorney during the litigation. We agree. We also agree with the county defendants, however, that where (as here) plaintiffs are the People of the State of California, acting through the Attorney General to enforce the laws of the state, an award of attorney fees under Code of Civil Procedure section 1021.5 in favor of the People is improper. Accordingly, we will reverse the award of attorney fees against the county defendants.

FACTUAL AND PROCEDURAL BACKGROUND

KAKE owns approximately 3,300 acres in Tehama County known as Burr Valley Estates. Since 1971, the property has been subject to a Williamson Act (Gov. Code, § 51200 et seq.) contract with the county.

“The Williamson Act establishes a mechanism for saving agricultural land by allowing counties to create agricultural preserves and then to enter into contracts with landowners within those preserves. (Gov. Code, § 51200 et seq.) A Williamson Act contract obligates the landowner to maintain the land as agricultural for 10 or more years, with resulting tax benefits. (Id., §§ 51240-51244.) Absent contrary action, each year the contract renews for an additional year, so that the use restrictions are always in place for the next nine to 10 years. (Id., § 51244.)” (Friends of East Willits Valley v. County of Mendocino (2002) 101 Cal.App.4th 191, 195 [ 123 Cal.Rptr.2d 708 ].)

In November .1998, KAKE applied for a lot line adjustment (LLA 98-46) with respect to Burr Valley Estates. At that time, former subdivision (d) of section 66412 (former section 66412(d)) provided that the Subdivision Map Act did not apply to “[a] lot line adjustment between two or more existing adjacent parcels, where the land taken from pne parcel is added to an adjacent parcel, and where a greater number of parcels than originally existed is not thereby created . . . .”

Tehama County Planning Director George Robson approved KAKE’s lot line adjustment after determining it did not create any additional parcels. The Tehama County Board of Supervisors did not review the lot line adjustment or make any findings prior to Robson’s approval.

Initially, documents were recorded showing 32 parcels resulting from the lot line adjustment. Ultimately, however,'amended documents were recorded in October 1999 showing 29 resulting parcels.

In May 2001, the Attorney General,' the Secretary of the State Resources Agency, and the Director of the Department of Conservation commenced this action on behalf of the People of the State of California (the People) by filing a complaint against the board of supervisors, Robson, and KAKE (jointly defendants), alleging violations of the Subdivision Map Act and the Williamson Act and seeking specific performance, injunctive relief, and declaratory relief. The People later joined the County of Tehama as a defendant.

Ultimately, the People filed a third amended complaint, that alleged three causes of action, only one of which (the second) is at issue here. The People alleged KAKE’s property consisted of only 24 parcels before the lot line adjustment. Thus, in the People’s view, the adjustment created a greater number of parcels than originally existed and was therefore subject to the Subdivision Map Act. The People alleged defendants had divided the property, in violation of the Subdivision Map Act, because, among other things, the board of supervisors did not “make the findings required by section 66474.4 specific to lands covered by Williamson Act contracts.” Elsewhere in the complaint, the People also asserted that even if LLA 98-46 did not create any additional parcels, the board of supervisors was required by “section 51257 of the Williamson Act [to make] certain specified findings . . . regarding agricultural compatibility because the property is enrolled in a Williamson Act contract.” The People sought an injunction to prevent the county from issuing development permits for any lots created by LLA 98-46 until the provisions of the Williamson Act and the Subdivision Map Act were met.

In May 2002, Robson filed a motion for summary adjudication against 'the State Resources Agency on the first and second causes of action. The motion was subsequently treated as having been filed on behalf of all the county defendants against all plaintiffs. The trial court heard that motion on September 27, 2002, and entered its order denying the motion on December 30, 2002.

Meanwhile, in October 2002, KAKE filed its own motion for summary judgment or summary adjudication .of the second .cause of action, KAKE argued there was no violation of the Subdivision Map Act because. LLA 98-46 did not result in the creation of any additional parcels. In KAKE’s view, the property consisted of at least 32 parcels before the lot line adjustment, and thus the adjustment actually reduced the number of parcels to 29. KAKE also argued that section 51257 of the Williamson Act does not apply to lot line adjustments within the boundaries of contracted lands where the resulting lots are at least 40 acres in size.

In January 2003, the People filed a cross-motion for summary adjudication of the second cause of action. The People contended findings were required under section 66474.4 of the Subdivision Map Act because LLA 98-46 did create additional parcels, and even if no additional parcels were created, findings were still required under section 51257 of the Williamson Act.

The court heard both motions in May' 2003, and on June 16, 2003, the court entered its order granting the People’s motion and denying KAKE’s motion. The court concluded that “at the time KAKE applied for a ‘lot line adjustment’ to create 29 parcels out of its 3,300 acres, the property then consisted of only two ... or, at most, four . . . preexisting parcels” and therefore it was necessary for KAKE to submit a parcel map to the board of supervisors and for “the Board to find that the proposed division is compatible with the Williamson Act.” In essence, the trial court concluded KAKE’s division of its property was subject to the Subdivision Map Act. The court also rejected KAKE’s argument that the action was barred by the statute of limitations in section 66499.37 of the Subdivision Map Act.

KAKE attempted to appeal from the order because it resolved all claims in the action against KAKE, but this court dismissed the appeal in December 2003 as being from a nonappealable order.

■ The trial court subsequently granted the People’s motion for summary adjudication of the third cause of action and the county defendants’ motion for judgment on the pleadings as to the first cause of action. Thereafter, in December 2004, the. court entered judgment consistent with its earlier rulings. In pertinent part, that judgment requires the county defendants “to make such findings as may be required under the Williamson Act and Subdivision Map Act in regard to subdivisions of five or more parcels” and prohibits KAKE “from offering for sale any of the lots created by LLA 98-46 until the provisions of the Williamson Act and the Subdivision Map Act pertaining to subdivisions of five or more parcels are met.”

The People then filed a motion for attorney fees under Code of Civil Procedure section 1021.5, seeking over $500,000 in fees from the county defendants. While that motion was pending, KAKE and the county defendants filed timely notices of appeal from the judgment. Subsequently, in May 2005, the court granted the People’s motion and ordered the county defendants to pay $173,450 in attorney fees. The county defendants filed a timely notice of appeal from that order, and the People filed a timely notice of cross-appeal.

DISCUSSION

I

KAKE’s and the County Defendants’ Appeals from the Judgment

A

Statute of Limitations

In its opposition to the People’s motion for summary adjudication, KAKE contended that to the extent the People’s second cause of action was premised on an alleged violation of the Subdivision Map Act, that cause of action was time-barred by the statute of limitations in section 66499.37. In pertinent part, that statute provides as follows: “Any action or proceeding to attack, review, set aside, void or annul the decision of an advisory agency, appeal board or legislative body concerning a subdivision, or of any of the proceedings, acts or determinations taken, done or made prior to such decision, or to determine the reasonableness, legality or validity of any condition attached thereto, shall not be maintained by any person unless such action or proceeding is commenced and service of summons effected within 90 days after the date of such decision. Thereafter all persons are barred from any such action or proceeding or any defense of invalidity or unreasonableness of such decision or of such proceedings, acts or determinations.”

KAKE contended the People’s claim of a violation of the Subdivision Map Act was time-barred by section 66499.37 because whatever the date of decision, it was no later than October 27, 1999, when the amended lot line adjustment was recorded, and the People did not commence their action until May 15, 2001, over one year and a half later.

The trial court concluded section 66499.37 did not apply because the People’s action was “not an action for administrative review to attack, annul, or review a decision of defendant Robson or any other of the County defendants.” KAKE contends the trial court erred in this conclusion.

We need not decide whether the trial court’s reasoning was correct because we conclude the statute does not apply for a different reason—specifically, KAKE has not shown that Robson (who approved LLA 98-46) qualified as “an advisory agency, appeal board or legislative body” within the meaning of the statute.

Implicitly conceding Robson did not qualify as an “appeal board or legislative body,” KAKE argues that he qualified as “an advisory agency” within the meaning of section 66415, which defines that term to mean “a designated official or an official body charged with the duty of making investigations and reports on the design and improvement of proposed divisions of real property, the imposing of requirements or conditions thereon, or having the authority by local ordinance to approve, conditionally approve or disapprove maps.” The problem is that KAKE makes no attempt to show how Robson fits within that definition. Specifically, KAKE points to nothing (either in the record or otherwise) to show that Robson: (1) was “charged with the duty of making investigations and reports on the design and improvement of proposed divisions of real property”; (2) was “charged with the duty of’ “imposing . . . requirements or conditions” “on the design and improvement of proposed divisions of real property”; or (3) “ha[d] the authority by local ordinance to approve, conditionally approve or disapprove maps.”

KAKE characterizes as “unfounded” the People’s “assertion that Director Robson did not have authority to approve or disapprove maps,” but this argument improperly attempts to shift the burden of proof and persuasion on this argument from KAKE to the People. It is not the People’s burden to show that Robson did not have the authority necessary to make him an “advisory agency” within the meaning of section 66415, but KAKE’s burden to show that he did. Because KAKE was the party attempting to rely on the statute of limitations as a defense to the People’s action, it was KAKE that had to show the existence of all elements necessary for Robson’s action to have triggered that statute. Similarly, because it is KAKE that now contends the trial court erred in finding the statute inapplicable, it is KAKE that bears the burden of persuading us the trial court erred in reaching that conclusion.

Unfortunately, the only effort KAKE makes to show Robson falls within section 66415 is an oblique argument that the “authority to approve a lot line adjustment” is the equivalent of the “authority to approve a map within the meaning of Government Code section 66415.” We are not persuaded. We understand the reference in section 66415 to the “the authority .. . to approve, conditionally approve or disapprove maps” to be a reference to the maps that are the subject of the Subdivision Map Act— tentative, final, and parcel maps. (See § 66425.) KAKE offers nothing to suggest Robson had the power to approve or disapprove such maps under any local ordinance. While he may have had the authority to approve a lot line adjustment, “[n]o tentative map, parcel map, or final map [can] be required as a condition to the approval of a lot line adjustment” (§ 66412), and therefore his authority to approve such an adjustment does not imply the authority to approve a map within the meaning of section 66415.

For the foregoing reasons, we conclude KAKE has failed to show the 90-day limitations period in section 66499.37 applies here.

B

Former Section 66412(d)

As we have noted, in 1998 when Robson approved LLA 98-46, former section 66412(d) provided that the Subdivision Map Act did not apply to “[a] lot line adjustment between two or more existing adjacent parcels, where the land taken from one parcel is added to an adjacent parcel, and where a greater number of parcels than originally existed is not thereby created . . . .” (Stats. 1994, ch. 458, § 2, p. 2575.) In 2001, the Legislature amended the statute to provide that this exception from the Subdivision Map Act applies only when the lot line adjustment is “between four or fewer existing adjoining parcels,” rather than “between two or more existing adjacent parcels” as the statute previously provided. (Stats. 2001, ch. 873, § 2.) In essence, the amendment limits this exception from the Subdivision Map Act to lot line adjustments involving two, three, or four adjacent parcels; lot line adjustments involving five or more adjacent parcels are subject to the Subdivision Map Act.

In the trial court, this case was litigated and resolved on the shared belief that the former version of section 66412(d) governed. On appeal, however, the People contend “[application of the current version of [the statute] is appropriate.” According to the People, this is so “because, on review of an injunction, the court applies the law in effect at the time of review.”

We need not decide whether application of the correct version of section 66412, subdivision (d) is appropriate because, as will be seen, even under the version of the statute that was in effect when Robson approved LLA 98-46, the People prevail.

C

Parcel Counting

The primary issue in these appeals is whether LLA 98-46 created more parcels than existed immediately before the lot line adjustment. Since it is undisputed LLA 98-46 resulted in 29 parcels, the dispositive question is whether there were 29 or more parcels before the adjustment. If there were 29 or more parcels, then LLA 98-46 did not create a greater number of parcels and KAKE’s division of its property was exempt from the Subdivision Map Act. If there were fewer than 29 parcels, then LLA 98-46 did create a greater number of parcels and the Subdivision Map Act applied.

All parties agree the question of how many parcels existed immediately before the lot line adjustment is a question of law to be answered based on the undisputed history of title to the land that now makes up Burr Valley Estates, as summarized and detailed by John Blodger, a land surveyor with the State Department of Conservation.

In determining how many parcels existed, it is helpful to note what is not disputed. It is undisputed that Burr Valley Estates is made up of all or part of seven different “sections,” numbered 4, 5, 9, 10, 15, 32, 33, and 34. (Further, the parcel count with regard to four of those sections—4, 9, 15, and 33—is undisputed; the parties agree there were 14 parcels in those sections: five in section 4, one in section 9, one in section 15, and seven in section 33.)

Where the parties disagree is on the number of parcels in sections 5, 10, 32, and 34. The People contend there were two parcels in sections 5 and 32, two parcels in section 10, and three parcels in section 34—which, when added to the 14 undisputed parcels, makes a total of 21 parcels. KAKE, on the other hand, contends there were four parcels in section 10, 14 parcels in section 34, and either five or eight parcels in sections 5 and 32 —for a total of either 37 or 40 parcels. The county defendants agree with KAKE that there were four parcels in section 10 and at least 10 parcels in section 34. According to them, however, since this means thqre were at least 28 parcels (14 plus four plus 10) before counting the parcels in sections 5 and 32, then even if the People are correct in counting only two parcels in those sections, the total still exceeds 29.

From the foregoing, it appears the crux of the case is the parcel-counting in sections 10 and 34. Thus, we begin our analysis with the number of parcels that were in sections 10 and 34 immediately before LLA 98-46.

Before turning to that analysis, we pause to note two principles the parties do not dispute. First, it is undisputed that mere common ownership of separate contiguous parcels does not result in merger of those parcels into a new, single parcel. (See § 66451.10, subd. (a) [“two or more contiguous parcels or units of land which have been created under the provisions of this division, or any prior law regulating the division of land, or a local ordinance enacted pursuant thereto, or which were not subject to those provisions at the time of their creation, shall not be deemed merged by virtue of the fact that the contiguous parcels or units are held by the same owner”].)

Second, it is undisputed that when two or more parcels that have been separately and distinctly described in an instrument of conveyance or security document are subsequently conveyed together using a single, consolidated legal description, those parcels are not merged into a single parcel absent an express statement by the grantor of the intent to do so in the instrument of conveyance. .This principle is based on Civil Code section 1093, which provides as follows: “Absent the express written statement of the grantor contained therein, the consolidation of separate and distinct legal descriptions of real property contained in one or more deeds, mortgages, patents, deeds of trust, contracts of sale, or other instruments of conveyance or security documents, into a subsequent single deed, mortgage, patent, deed of trust, contract of sale, or other instrument of conveyance or security document (whether by means of an individual listing of the legal descriptions in a subsequent single instrument of conveyance or security document, or by means of a consolidated legal description comprised of more than one previously separate and distinct legal description), does not operate in any manner to alter or affect the separate and distinct nature of the real property so described in the subsequent single instrument of conveyance or security document containing either the listing of or the consolidated legal description of the parcels so conveyed or secured thereby.”

Civil Code section 1093 was not enacted until 1985 (Stats. 1985, ch. 911, § 1, p. 2905), long after the conveyances at issue here; however, the Legislature declared that the statute did “not constitute a change in, but is declaratory of, the existing law.” (Civ. Code, § 1093.) No party disputes the accuracy of this declaration. As will be seen, while we have reason to question the declaration’s accuracy, there is no need here to actually deter mine whether section 1093 accurately expresses the law as it was before 1986. Accordingly, we will assume for purposes of this.case that it does.

1. Section 10

With the foregoing principles in mind, we turn to the counting of parcels in section 10. As relevant here, that section was originally comprised of four separate parcels created by federal patents. By 1904, Charles Hesse owned two of those parcels: one consisting of the west half of the west half of the section and the other consisting of the east half of the southwest quarter and the southeast quarter of the northwest quarter of the section. The other two parcels—one consisting of the south half of the northeast quarter and the west half of the southeast quarter of the section, and the other consisting of the east half of the southeast quarter of the section—were owned by G. C. Garrett and C. E. Tinkham. (See appen., post, at p. 458.)

A county road known as Ridge Road ran through all four parcels roughly east to west in the south half of the section. In 1904, a land exchange accomplished by three different deeds conveyed all of the land in section 10 south of Ridge Road to B. A. Bell, while Charles Hesse became the owner of all the land in section 10 north of Ridge Road that had been part of the four parcels. (See appen., post, at p. 458.) The land south of Ridge Road is not part of Burr Valley 'Estates and is not at issue here. What is at issue is how many parcels Charles Hesse owned in section 10 following the 1904 land exchange.

The first deed in the 1904 land exchange was executed on January 7, 1904. In that deed, Garrett and Tinkham conveyed the parts of the two parcels they owned in section 10 that lay north of Ridge Road to Charles Hesse by a single metes and bounds description that did not refer to the two existing parcels from which the land was taken.

The parties disagree about the number of parcels conveyed in this deed. KAKE and the county defendants contend the deed conveyed two new parcels to Charles Hesse—specifically, those parts of the two parcels Garrett and Tinkham owned in section 10 that lay north of Ridge Road (we will refer to these as fractional parcels). The People, on the other hand, contend the deed conveyed only a single new parcel—the parcel described in the deed. For the reasons that follow, we agree with the People.

“In the construction of boundaries, the intention of the parties is the controlling consideration. [Citation.] Whenever possible, a court should place itself in the position of the parties and ascertain their intent, as in the case of any contract. As stated in Miller & Lux, Inc., v. Secara, 193 Cal. 755 [ 227 Pac. 171 ], ‘Intention, whether express or shown by surrounding circumstances, is all controlling ....’” (Machado v. Title Guarantee and T. Co. (1940) 15 Cal.2d 180, 186 [ 99 P.2d 245 ].)

“[E]xtrinsic evidence is always admissible to explain the calls of a deed for the purpose of their application to the subject-matter, and thus to give effect to the deed. [Citation.] In construing a doubtful description in a grant, the court must assume as nearly as possible the position of the contracting parties, and consider the circumstances of the transaction between them, and then read and interpret the words used in the light of these circumstances.” (Thompson v. Motor Road Co. (1890) 82 Cal. 497, 500 [ 23 P. 130 ].)

Here, no party points to any evidence of surrounding circumstances that would assist in divining the intent of the parties to the various transactions at issue, including the 1904 deed from Garrett and Tinkham to Charles Hesse. As for the deed itself, there is nothing in it suggesting Garrett and Tinkham intended to convey two new fractional parcels to Charles Hesse, as opposed to a single new parcel. The deed contains a single metes and bounds description of what by all appearances is one tract of land. As the People point out, the deed does not make “any reference to the old patent parcel boundary that . . . divide[d] the original holdings of Tinkham and Garrett in section 10.” Furthermore, the deed specifies that what is being conveyed is “all that certain lot and parcel of land” which the deed then goes on to describe. (Italics added.) Before it was filled out, that part of the preprinted deed looked like this: “all th_certain lot_and parcel_of land.” Had Garrett and Tinkham intended to convey two parcels instead of one, they could have filled out the deed to convey “all those certain lots and parcels of land” thereafter described, but they did not do so. Instead, they filled the blank after the “th” with “at” to create the word “that” and put small lines through the blanks after “lot” and “parcel” showing that they intended to use the singular of those words, rather than the plural. Thus, by all appearances, the deed conveyed a single new parcel created from parts of the two parcels Garrett and Tinkham owned in section 10, rather than two new parcels, each created from one of those parts.

■ Both KAKE and the county defendants suggest we should not be so quick to infer an intent on the part of Garrett and Tinkham and Charles Hesse to (as the county defendants put it) “destroy old parcel boundaries and entirely subsume them within the new ones” because the combination of parts of existing parcels into a single new parcel would make the land involved less valuable. On this point, KAKE observes that “[m]erger [of the parts of the existing parcels] would result in a loss of the landowner’s right later to divide the property without complying with then-applicable subdivision laws.” While that is certainly true, if we look to the state of the law in Tehama County in 1904, we find no reason that Charles Hesse would have been concerned about later difficulties in dividing the property he received from Garrett and Tinkham.

California’s first subdivision map statute was enacted in 1893 (Stats. 1893, ch. 80, § 1, p. 96; see Morehart v. County of Santa Barbara (1994) 7 Cal.4th 725, 761 [ 29 Cal.Rptr.2d 804 , 872 P.2d 143 ]), but it was not until the 1970’s that divisions of land in which only four or fewer parcels were created (minor subdivisions) were subject to statewide regulation. (See van’t Rood v. County of Santa Clara (2003) 113 Cal.App.4th 549, 565-566 [ 6 Cal.Rptr.3d 746 ].) Although such divisions of land “still could be regulated by local ordinance” (Stell v. Jay Hales Development Co. (1992) 11 Cal.App.4th 1214, 1224 [ 15 Cal.Rptr.2d 220 ]), Tehama County did not have an ordinance regulating minor subdivisions before 1972. What this means is that in 1904 there would have been no reason for Charles Hesse to be concerned that Garrett and Tinkham were conveying only one parcel to him, because the law at the time allowed him to divide that parcel into two, three, or even four parts and sell those parts separately without any restriction. No party offers any reason why Charles Hesse should have anticipated a change in. the law that would have restricted his right to divide his property. Given this circumstance, we find no reason to believe Garrett and Tinkham intended to convey two new fractional parcels to Charles Hesse rather than the single new parcel described in the deed.

KAKE attempts to avoid this conclusion by relying on the antimerger rule in Civil Code section 1093. KAKE contends that finding an intent to “ ‘obliterate the previous patent boundaries’ ” “from use of the metes and bounds deed[] would make Civil Code section 1093 irrelevant,” and under Civil Code section 1093 “[t]he use of a metes and bounds deed does not result in merger of jointly described parcels.”

KAKE reads the statute too broadly. As the People point out, Civil Code section 1093 “concerns only those situations where one party is conveying entire pre-existing parcels in a consolidated description” and “does not apply to situations where portions of pre-existing parcels, never before described separately, are being conveyed.” KAKE contends this is a “creative interpretation of the language of [Civil Code] section 1093” that “is not supported by citation to any authority.” KAKE is correct on the latter point, but no citation, to authority is necessary when, as here, the statutory language is clear. (See DaFonte v. Up-Right, Inc. (1992) 2 Cal.4th 593, 601 [ 7 Cal.Rptr.2d 238 , 828 P.2d 140 ] [when statutory language is clear and unambiguous, there is no need for construction].) Undeniably, by its plain terms, Civil Code section 1093 applies only when two or more “separate and distinct legal descriptions of real property contained in one or more [previous] instruments of conveyance or security documents” are later consolidated in “a subsequent. . . instrument of conveyance or security document,” either by “an individual listing of the legal descriptions” or “a consolidated legal description comprised of more than one previously separate and distinct legal description.”

Here, no “separate and distinct legal description” of either of the fractional parcels KAKE contends Garrett and Tinkham conveyed to Charles Hesse is contained in any previous instrument of conveyance or security document. It necessarily follows then that the 1904 deed from Garrett and Tinkham to Charles Hesse does not contain “an individual listing of [any previously separate and distinct] legal descriptions” of those fractional parcels or “a consolidated legal description comprised of more than one previously separate and distinct legal description” of those fractional parcels. Under these circumstances, the rule stated in Civil Code section 1093 simply does not apply.

The county defendants contend that even if Civil Code section 1093 does not directly apply “to transfers of fractions of separate parcels, it is certainly a persuasive determination, in an analogous setting, that the use of a single legal description does not evince an intention to merge heretofore separate pieces of land.” Thus, even though Civil Code section 1093 does not apply here,, the county defendants would have us reach the same result that Civil Code section 1093 would compel if it did apply because the statute is “presumably representative of a consistent body of law with consistent purpose and effect.”

This is a rather remarkable contention. In effect, without citing any legal authority in support of their position, the county defendants would have us conclude that California common law provides (and has long provided) an antimerger rule identical to the one set forth in Civil Code section 1093, except that (unlike Civil Code section 1093) it applies to fractional parcels that have never before been described separately in any security document or instrument of conveyance. According to the county defendants, such a rule must exist (even though we have been given no evidence of it) because otherwise “major landowners . . . would routinely surrender their successor’s ability to separately use and sell portions of their land, to their potentially severe financial detriment.”

We are not inclined to create a rule of common law from whole cloth and apply it to transactions that occurred more than 100 years ago. Moreover, we are particularly reluctant to extend the rule stated in Civil Code section 1093 to situations beyond the scope of the statutory language because, notwithstanding the Legislature’s assertion in the statute, we have reason to question whether the rule stated in the statute actually existed before the statute was enacted. No one has cited, nor have we found, any authority (other than Civil Code section 1093 itself) supporting the assertion that before 1986, a consolidated legal description used in a deed to convey two or more parcels that had previously been separately and distinctly described in one or more instruments of conveyance or security documents could not be deemed to merge, the preexisting parcels absent an express written statement of the grantor contained in the later deed. In other words, we have found no evidence that.the rule stated in Civil Code section 1093 was the law before the statute was enacted. As far as we can determine, historically the law that governs the interpretation of deeds has been that “the intention of the parties is the controlling consideration” (Machado v. Title Guarantee and T. Co., supra, 15 Cal.2d at p. 186 ), and “[i]n construing a doubtful description in a grant, the court must assume as nearly as possible the position of the contracting parties, and consider the circumstances of the transaction between them, and then read and interpret the words used in the light of these circumstances” (Thompson v. Motor Road Co., supra, 82 Cal. at p. 501 ). This law did not make an express statement of merger by the grantor an absolute prerequisite to finding an intent to merge two or more previously separate parcels.

The Legislature’s assertion that a statute is declaratory of existing law does not make it so, if the law found in the code books and case books does not in fact support that assertion. Nevertheless, because the validity of Civil Code section 1093, as it applies to the situations specifically described in the statute, has no bearing on the outcome of this case (as will be shown), we need not actually determine whether the Legislature’s assertion in the statute was accurate. We do, however, conclude the apparent absence of any law consistent with the antimerger rule in Civil Code section 1093 prior to the statute’s enactment is further reason for us to decline the county defendants’ request that we recognize an analogous antimerger rule that applies to fractional parcels that have never before been separately and distinctly described, and that we apply that rule to transactions that occurred more than 100 years ago.

Based on the foregoing reasoning, we conclude that in 1904 when Garrett and Tinkham conveyed to Charles Hesse that portion of their property in section 10 lying north of Ridge Road using a single metes and bounds description and identifying the property conveyed as “that certain lot and parcel of land,” they intended to convey to Charles Hesse the single parcel they described in the deed to him and not two fractional parcels that were not then (and have never since been) separately described.

The foregoing conclusion disposes of the property in the eastern half of section 10 that is included in Burr Valley Estates—it consists of one parcel, not two. This takes us to the second deed in the 1904 land exchange and the property in the western half of section 10. On January 12, 1904, Charles Hesse conveyed to B. A. Bell the parts of the two parcels he owned in the west half of section 10 lying south of Ridge Road. We are not concerned here with the property Charles Hesse conveyed to Bell, as that property—like the other property in section 10 south of Ridge Road—is not part of Burr Valley Estates and therefore not at issue here. What is at issue here is the property that Charles Hesse retained.

We conclude the property Charles Hesse retained consisted of two distinct parcels—specifically, the remainders of the two parcels off of which he carved the land south of Ridge Road -to give to Bell. Indeed, the People do not argue otherwise. Instead, they contend “the facts of this case make th[e] inquiry [into whether one or two parcels remained] irrelevant” because “[i]n every case where a landowner conveyed out fractions of pre-existing parcels, the landowner later conveyed out the remainder fractions as well.” According to the People, the later conveyance has the same legal consequences as the 1904 conveyance from Garrett and Tinkham to Charles Hesse—even if Charles Hesse retained two parcels instead of one, his later use of a single legal description to convey those two previously undescribed parcels, without reference to the preexisting patent parcel lines, demonstrates an intent to merge those parcels into one.

We turn to Charles Hesse’s disposition of what remained in the west half of section 10 after his 1904 conveyance to Bell to determine if that is so. After Charles Hesse conveyed the land south of Ridge Road to Bell, he was left with the remainder of each of the two parcels north of the road. In 1918, he and his wife conveyed that land, along with other land, to Albert Montgomery in a deed that described the land as “all those certain lots and parcels of land, situate, lying and being in the County of Tehama, State of California, and bounded and particularly described as follows, to-wit: All of fractional Section Four (4), all of Section Nine (9), the West half of the Northwest quarter (W1/2 of NW 1/4), the Southeast quarter of the Northwest quarter (SE 1/4 of NW 1/4), the North half of the Southwest quarter (N1/2 of SW 1/4), and all that part of the South half of the Southwest quarter (S 1/2 of SW 1/4), lying North of the Ridge Road in Section Ten (10). All that part of the Northwest quarter of the Northwest quarter (NW 1/4 of NW 1/4) lying North of the Ridge Road, in Section Fifteen (15), all in Township Twenty-six (26), North of Range Five (5) West, M.D.M., and containing 1698 acres, more or less.” (Italics added.) (See appen., post, at p. 459.)

The italicized portion of the legal description above describes the property in section 10 north of Ridge Road that Charles Hesse conveyed to Montgomery. The People’s position is that because this “description does not retain any reference to the old patent parcel boundary which used to lie within the remainder parcel,” “the remainder parcel that was conveyed to Montgomery [w]as just one parcel.” KAKE counts this property as two parcels because there is no “language of merger” in the deed, and therefore under Civil Code section 1093, the two preexisting parcels were not merged by the conveyance to Montgomery. Civil Code section 1093 does not apply, however, because the two parcels at issue—what remained after Charles Hesse conveyed his land in section 10 south of Ridge Road to Bell—were never separately and distinctly described in any instrument of conveyance or security document. Moreover, as we have explained already, we have found no basis for recognizing the existence of a rule like that in Civil Code section 1093 that applies to fractional parcels that have not been so described. Accordingly, we are left with nothing more than the description in the deed to ascertain whether Charles Hesse intended to convey the undescribed fractional parcels or the one parcel actually described in the deed.

We conclude the latter answer is correct. Although this deed (unlike the 1904 deed discussed above) refers to “all those certain lots and parcels of land,” the use of the plural is explained by the fact that in addition to conveying the land in section 10 to Montgomery, Charles Hesse also conveyed to him in this deed various other parcels—specifically, parcels in section 4, section 9, and section 15. Under these circumstances, absent anything in the deed or any extrinsic evidence suggesting Charles Hesse intended to maintain the parcel boundary that previously separated the two fractional parcels he retained in the west half of section 10 after conveying the land south of Ridge Road to Bell, we conclude the property in the west half of section 10 that is now part of Burr Valley Estates consists of one parcel, not two.

2. Section 34

We turn now to section 34. That section was originally comprised of eight separate parcels created by federal patents. By 1907, Charles Hesse owned all eight parcels, including two parcels comprising the southeast quarter of the section—namely, the north half of the quarter and the south half of the quarter. In 1908, he conveyed to Frances Hesse 100 acres in the southeast quarter of the section, which the deed described by metes and bounds without reference to the two existing parcels from which the land was taken. Essentially, what was conveyed were portions of the existing parcels lying: (1) south of a road (Johnson Road) running roughly east to west near the northern boundary of the quarter; and (2) east of a line running from the road to the south boundary of the quarter. (See appen., post, at p. 460.)

The parties disagree about the number of parcels conveyed in this deed. KAKE contends that in making this conveyance, Charles Hesse “cut pieces out of [the two existing] parcels . . . , creating two new legal parcels,” which he conveyed to Frances Hesse. The county defendants agree. The People, on the other hand, contend Charles Hesse conveyed only a single new parcel to Frances Hesse—the parcel described in the deed.

For the reasons set forth above in connection with our analysis of section 10, we agree with the People. The deed to Frances Hesse described a single “lot and parcel of land” without reference to the boundary between the preexisting parcels from which the new parcel was taken. Absent anything in the deed or any extrinsic evidence suggesting Charles Hesse intended to maintain that parcel boundary, we conclude Charles Hesse intended to convey the single parcel he described in the deed to Frances Hesse, and not two fractional parcels that were not then (and have never since been) separately described.

In 1910, Charles Hesse conveyed another 100 acres south of Johnson Road to Frances Hesse, again using a metes and bounds description without reference to the existing parcels from which the land was taken. (See appen., post, at p. 460.) KAKE contends this deed created nine parcels where there were five before. For the reasons set forth above, KAKE is mistaken. Like the 1908 deed, the 1910 deed to Frances Hesse described a single “lot and parcel of land” without reference to the boundaries between the preexisting parcels from which the new parcel was taken. Absent anything in the deed or any extrinsic evidence suggesting Charles Hesse intended to maintain those parcel boundaries, we conclude Charles Hesse intended to convey the single parcel he described in the deed to Frances Hesse, and not five new fractional parcels that were not then (and have never since been) separately described.

That brings us to what happened to section 34 after 1910. Following his two deeds to Frances Hesse, Charles Hesse continued to own three of the original patent parcels in that section and the fractional remainders of four others. In 1913, Charles Hesse conveyed all of that land, along with other land in section 33, to R. L. Douglas by conveying “all those certain lots and parcel [sic] of land, situate, lying and being in the County of Tehama, State of California, and bounded and particularly described as follows, to wit: [f] All of Section thirty three (33) in Township twenty seven (27) North Range Five (5) West M. D. M. and all of Section thirty four (34) of said Township, and Range, except Two Hundred (200) acres thereof heretofore deeded by party of the first part to Frances Hesse.” (Italics added.) (See appen., post, at p. 461.)

KAKE contends that the italicized language in this deed conveyed seven parcels to Douglas—the three original patent parcels and the four fractional parcels that Charles Hesse owned after his 1910 deed to Frances Hesse. The People contend the italicized language conveyed only one parcel to Douglas because it merged all of Charles Hesse’s remaining land in section 34 into one new parcel, even though that land included “the entirety of three pre-existing patent parcels.” (Italics omitted.) Recognizing that this position presents a potential conflict with the antimerger rule in Civil Code section 1093, the People assert that “even if the boundaries of the three preexisting parcels wholly contained in the conveyance should today be honored, that would only add three to the State’s parcel count,” raising it to 24, which still means “the County’s lot line adjustment for KAKE [was] illegal.” (Italics omitted.)

We need not decide which position is correct, however, because even if KAKE is correct in asserting that the 1913 deed conveyed seven parcels to Douglas, it makes no difference in the outcome of the case. Based on our previous analysis, accepting KAKE’s position on the 1913 deed would result in only nine parcels in section 34—the two conveyed in 1908 and 1910 to Frances Hesse and the seven conveyed in 1913 to Douglas. As we will explain, nine parcels in section 34 is not enough to bring the total number of parcels in Burr Valley Estates to 29 before the lot line adjustment, which is a prerequisite for KAKE and the county defendants to prevail on their appeals from the judgment.

3. Sections 5 and 32

The reason for this result lies in the parcel-counting in sections 5 and 32. As noted above, the People contend there were two parcels in sections 5 and 32, while KAKE contends there were either five or eight parcels in those sections. Again, we agree with the People.

The history of the lands in sections 5 and 32 is complicated. For our-purposes, however, we need only focus on a portion of that history. By 1936, all of the land in sections 5 and 32 (along with other land in sections 6 and 31) was owned by the estate of Mandus Johnson. (See appen., post, at p. 462.) In April 1936, the executors of the estate conveyed all of that land to Andrew Jr., Silas, Olaf, and Malen Johnson. By 1944, following a series of conveyances, Andrew Jr. owned an undivided one-quarter interest in the property and Silas owned an undivided three-quarters interest in the property. (See appen., post, at p. 463.) Then, in 1948, in an action filed by the executors of Andrew Jr.’s estate against Silas and others, the Tehama County Superior Court entered an order partitioning the property (known as Burr Valley Ranch) between Silas and his wife and Andrew Jr.’s heirs. As the People point out, “Insofar as is pertinent here, Silas Johnson obtained a parcel consisting essentially of the west half of section 5, while [Andrew Jr.’s heirs] obtained a parcel that consisted essentially of the east half of section 5 and a part of the east half of section 32.” (See appen., post, at p. 463.) The order described the property given to Andrew Jr.’s heirs as “the following portion of Burr Valley Ranch,” followed by a metes and bounds description that did not reference any preexisting patent parcel lines, then described the property given to Silas and his wife as “the remaining portion of said Burr Valley Ranch,” likewise followed by a metes and bounds description that did not reference any preexisting patent parcel lines. (Italics added.) The order also noted that various “issues and questions” were reserved for future determination, including “[w]hether the Court shall determine that the owners of the respective portions of the Burr Valley Ranch, as partitioned and divided herein, shall bear equally the costs of a division fence between said two parcels as herein partitioned when such fence is constructed.” (Italics added.)

The People contend the partition decree “created two new parcels that are entirely independent of any prior parcelization in the two relevant sections.” KAKE, on the other hand, contends the partition resulted in four parcels in section -5 because of the continued existence of parts of the original patent parcel boundaries. Again, however, KAKE’s contention relies on Civil Code section 1093, but the rule in Civil Code section 1093 does not apply here because the property at issue consisted only of fractional parcels, and we have found no basis to recognize the existence of a rule analogous to the one found in Civil Code section 1093 that applies to fractional parcels. Under these circumstances, we agree with the People that before the lot line adjustment sections 5 and 32 contained two preexisting parcels, rather than the five to eight parcels KAKE advocates.

It is because of this result that we need not decide whether the 1913 deed from Charles Hesse to Douglas of land in section 34 conveyed one, four, or seven parcels. Even if it conveyed the greater number, the total number of parcels in Burr Valley Estates immediately before LLA 98-46 would have been only 27—five in section 4, one in section 9, one in section 15, seven in section 33, two in section 10, two in sections 5 and 32, and nine in section 34 (seven conveyed in 1913, plus the two parcels conveyed to Frances Hesse). Since this is fewer than the 29 parcels created by the lot line adjustment, the trial court correctly concluded the adjustment was not exempt from the Subdivision Map Act. Accordingly, we will affirm the judgment.

II

The County Defendants’ and the People’s Appeals from the Attorney Fees Award

The county defendants challenge the award of attorney fees against them on several bases. First, they contend the private attorney general statute (Code Civ. Proc., § 1021.5) does not authorize an award of attorney fees to the Attorney General under the circumstances of this case, if ever. Second, they contend that even if the Attorney General can, in some circumstances, qualify for a fee award under this statute, the People’s “personal interests” in this particular litigation make a fee award inappropriate here. Finally, the county defendants contend the trial court abused its discretion in basing the fee award on perceived misconduct by their former attorney rather than on the statutory criteria.

For their part, the People challenge the fee award as inadequate. They agree with the county defendants that “the trial court’s sua sponte imposition of attorney’s fees as sanctions for the vexatious behavior of the County was procedurally improper,” but contend Code of Civil Procedure section 1021.5 “authorize^] a full award of attorney’s fees to the State in this case.” They contend we should reverse the fee award and remand the case to the trial court to reconsider the matter.

A

The “Obdurate Behavior” Rule

We begin with the issue on which both sides agree—whether the trial court’s rationale for awarding attorney fees under Code of Civil Procedure section 1021.5 was flawed. To answer that question, we turn to the language of the statute, which in relevant part provides as follows: “Upon motion, a court may award attorneys’ fees to a successful party against one or more opposing parties in any action which has resulted in the enforcement of an important right affecting the public interest if: (a) a significant benefit, whether pecuniary or nonpecuniary, has been conferred on the general public or a large class of persons, (b) the necessity and financial burden of private enforcement, or of enforcement by one public entity against another public entity, are such as to make the award appropriate, and (c) such fees should not in the interest of justice be paid out of the recovery, if any. With respect to actions involving public entities, this section applies to allowances against, but not in favor of, public entities . . . unless one or more successful parties and one or more opposing parties are public entities . . . .”

“[T]he Legislature adopted [Code of Civil Procedure] section 1021.5 as a codification of the ‘private attorney general’ attorney fee doctrine that had been developed in numerous prior judicial decisions.” (Woodland Hills Residents Assn., Inc. v. City Council (1979) 23 Cal.3d 917, 933 [ 154 Cal.Rptr. 503 , 593 P.2d 200 ].) “Entitlement to fees under [Code of Civil Procedure] section 1021.5 requires a showing that the litigation: ‘(1) served to vindicate an important public right; (2) conferred a significant benefit on the general public of a large class of persons; and (3) imposed a financial burden on plaintiffs which was out of proportion to their individual stake in the matter.’ ” (California Licensed Foresters Assn. v. State Bd. of Forestry (1994) 30 Cal.App.4th 562, 569 [ 35 Cal.Rptr.2d 396 ], quoting Baggett v. Gates (1982) 32 Cal.3d 128, 142 [ 185 Cal.Rptr. 232 , 649 P.2d 874 ].)

Purporting to apply the statutory criteria, the trial court concluded “the ordinary costs of litigation that the [People] have incurred here did not confer a benefit on the public that transcends their interest in the controversy, that is out of proportion to the [People]’s interest in this matter. No doubt, the public interest of the state at large has been well-served by this litigation; but serving the interests of the state at large is the public responsibility of these plaintiffs____”

Despite this conclusion, the court went on to examine what it believed was another “aspect” of the private attorney general doctrine—“the ‘obdurate behavior’ rule.” In the trial court’s assessment, where a plaintiff has succeeded in conferring a substantial benefit on the general public, it is “appropriate” under Code of Civil Procedure section 1021.5 “to pass on that portion of the attorney fees incurred for the ‘financial burden’ that was created by the obdurate behavior” of the defendants. Applying this “aspect” of Code of Civil Procedure section 1021.5, the trial court concluded the People should recover $173,450 in attorney fees because “the manner in which the litigation was conducted by the County defendants caused the time and expense of this litigation to be expanded enormously.” According to the trial court, the services of one deputy attorney general and one paralegal were entirely attributable to “the unnecessary and vexatious procedures pursued by” the county defendants, and the amount of time the lead deputy attorney general spent on the case was “increased by at least a third as a result of the vexatious and obdurate . . . tactics that were used.” Thus, the court concluded “the necessity and financial burden of these expenses are such as to make an award of these fees appropriate.”

We agree with the People and the county defendants that the trial court’s rationale for awarding $173,450 in fees under Code of Civil Procedure section 1021.5. was erroneous. Contrary to the trial court’s conclusion, that statute does not allow an award of fees based on a defendant’s “obdurate behavior” during the litigation.

The trial court’s error in this regard stemmed from a misreading of this court’s opinion in County of Inyo v. City of Los Angeles (1978) 78 Cal.App.3d 82 [ 144 Cal.Rptr. 71 ]. In that case, Inyo County moved for an award of attorney fees against the City of Los Angeles in a writ proceeding in which the county had succeeded in requiring the city to prepare an environmental impact report covering the city’s increased extraction and use of Owens Valley groundwater. (Id. at pp. 84-85.) This court noted “the general American doctrine which denies attorney fees to victorious litigants unless provided by statute or contract,” then observed that “Inyo County’s fee application is grounded on three rules or theories which various courts have recognized as nonstatutory exceptions to the general doctrine: the private attorney general rule, the substantial benefit rule and the vexatious litigant rule. (The county terms the last the ‘obdurate behavior’ rule.)” (Id. at p. 86.) The court went on to deny the county’s fee request under each of those three doctrines. (Id. at pp. 89-93.)

Notably, in denying a fee award under the vexatious-litigant obdurate-behavior rule, this court did not actually determine whether it had the power to make such an award; instead, the court “assume[d] existence of power to make the award on this ground.” (County of Inyo v. City of Los Angeles, supra, 78 Cal.App.3d at p. 91 .) Nine months later, however, in Bauguess v. Paine (1978) 22 Cal.3d 626 [ 150 Cal.Rptr. 461 , 586 P.2d 942 ], the California Supreme Court concluded trial courts did not have the inherent power to impose sanctions in the form of attorney fees for alleged misconduct. Soon thereafter, the Legislature gave trial courts that power by enacting section 128.5 of the Code of Civil Procedure. (See City of Long Beach v. Bozek (1982) 31 Cal.3d 527, 537 [ 183 Cal.Rptr. 86 , 645 P.2d 137 ].)

From the foregoing (if not from the language of the statute itself), it is apparent Code of Civil Procedure section 1021.5 does not provide any basis for a court to award attorney fees for vexatious litigation or obdurate behavior. Code of Civil Procedure section 1021.5 is a codification of the private attorney general doctrine, which is an entirely distinct basis for recovering, attorney fees from,the sanctions-based doctrine that now finds expression in various other statutes, such as section 128.5 of the Code of Civil Procedure.

Here, the People did not seek an award of attorney fees under section 128.5 of the Code of Civil Procedure, or any other statute for that matter, based on the conduct of the county defendants in' litigating the case; they sought an award only under the private attorney general doctrine codified in Code of Civil Procedure section 1021.5. Accordingly, the trial court erred in not limiting itself to a consideration of the criteria in that statute.

B

The Necessity Criterion

Having concluded the trial court erred in its application of Code of Civil Procedure section 1021.5 to this case, we turn to the arguments of the county defendants that under no circumstances could an award under that statute be appropriate here.

Consistent with the moniker of the doctrine it codified (the private attorney general doctrine), Code of Civil Procedure section 1021.5 originally precluded public entities from receiving fees under the statute. (See, e.g., City of Carmel-by-the-Sea, v. Board of Supervisors (1986) 183 Cal.App.3d 229, 254-256 [227 Cal-Rptr. 899].) In 1993, however, the Legislature amended the statute to its present form, which allows a .public entity to recover attorney fees from another public entity. (Stats. 1993, ch. 645, § 2, p. 3747.)

Here, the People, acting through the Attorney General, the Secretary of the State Resources Agency, and the Director of the State Department of Conservation, recovered fees from the County of Tehama after prevailing in an action to enforce the Subdivision Map Act and the Williamson Act. At first glance, at least, this situation appears to be covered by the 1993 amendment to Code of Civil Procedure section 1021.5. The county defendants argue, however, that the fee award here was not proper because this case is a quintessential “public enforcement” case to which Code of Civil Procedure section 1021.5 has never applied. According to them, the case law interpreting Code of Civil Procedure section 1021.5 has always recognized a “distinction between the ‘private enforcement’ that is rewarded by the private attorney general doctrine and the ‘public enforcement’ whose absence justifies the doctrine’s existence.” In their view, where “public enforcement” is available, fees under Code of Civil Procedure section 1021.5 are not, and since this case constitutes quintessential “public enforcement,” Code of Civil Procedure section 1021.5 fees are not available here.

The county defendants draw their distinction between “private enforcement” and “public enforcement” from case law predating the 1993 amendment to the statute. At the time, fees were available under Code of Civil Procedure former section 1021.5 only where “the necessity ... of private enforcement . . . [was] such as to make the award appropriate.” In City of Sacramento v. Drew (1989) 207 Cal.App.3d 1287 [ 255 Cal.Rptr. 704 ], this court explained that “the statutory requirement of ‘necessity ... of private enforcement’ addresses the issue of the comparative availability of public enforcement.” (Id. at p. 1298.) The court went on to observe that “if there is a public attorney general available to enforce the important right at issue there is no utility in inducing a private attorney general to duplicate the function.” (Id. at p. 1299.)

Based on the foregoing, the county defendants suggest that recovery of fees under Code of Civil Procedure section 1021.5 is not available when (as here) the Attorney General is available and chooses to act, because such action is by its very nature “public enforcement,” the availability of which renders “private enforcement”—the type of enforcement to which the statute applies—unnecessary and thus unqualified for a fee award under the statute.

This argument suffers two fatal flaws. First, the necessity criterion in Code of Civil Procedure section 1021.5 has never been interpreted that strictly. Indeed, even where a private party “colitigates with a governmental entity on behalf of the public or a large class of persons, whether by way of intervention or ... by consolidation of separately filed cases,” fees under Code of Civil Procedure section 1021.5 may be available if “the colitigating private party rendered necessary and significant services of value to the public or to a large class of persons benefited by the result of the litigation.” (Committee to Defend Reproductive Rights v. A Free Pregnancy Center (1991) 229 Cal.App.3d 633, 642 [ 280 Cal.Rptr. 329 ].) Thus, even before the 1993 amendment, neither the availability nor the actuality of “public enforcement”—that is, enforcement by a public entity—necessarily precluded a fee award under Code of Civil Procedure section 1021.5.

Second, even if we were to assume that the county defendants’ analysis was valid before the 1993 amendment, it does not withstand scrutiny following that, amendment. Under the 1993 amendment, the availability of enforcement by a public entity (i.e., “public enforcement”) has no bearing on the necessity criterion in an action involving public entities on both sides. In such a case, the court can award fees under the statute if (among other things) “the necessity ... of enforcement by one public entity against another public entity, [is] such as to make the award appropriate.” The question in such a case is not whether private enforcement was necessary but whether public enforcement—that is, enforcement by one public entity against another—was necessary. Obviously, in such a case the availability of “public enforcement” cannot preclude a fee award, or no public entity would ever be able to recover fees under Code of Civil Procedure section 1021.5, in derogation of the very terms of the 1993 amendment to the statute.

In summary, we do not agree with the county defendants that “the sovereign’s deliberate legal action, through its chief law officer, to enforce its own laws” can never qualify for a fee award under Code of Civil Procedure section 1021.5 because such action constitutes “public enforcement” to which the statute does not apply. The distinction between “public enforcement” and “private enforcement” the county defendants rely upon has no bearing in an action, like this one, involving public entities on both sides.

C

The Financial Burden Criterion

That does not resolve the matter, however, because the county defendants also rely on the financial burden criterion in the statute as a basis for their argument that an award of attorney fees under Code of Civil Procedure section 1021.5 can never be made to the Attorney General. As noted, Code of Civil Procedure section 1021.5 allows a fee award where (among other things) “the necessity and financial burden of private enforcement, or of enforcement by one public entity against another public entity, are such as to make the award appropriate.” Traditionally, the financial burden criterion has been deemed satisfied “ ‘when the cost of the claimant’s legal victory transcends his personal interest, that is, when the necessity for pursuing the lawsuit placed a burden on the plaintiff “out of proportion to his individual stake in the matter.” ’ ” (Woodland Hills Residents Assn., Inc. v. City Council, supra, 23 Cal.3d at p. 941 , quoting County of Inyo v. City of Los Angeles, supra, 78 Cal.App.3d at p. 89 ; see also Serrano v. Priest (1977) 20 Cal.3d 25, 46, fn. 18 [ 141 Cal.Rptr. 315 , 569 P.2d 1303 ].)

The county defendants contend that “[w]here the successful party is the Attorney General representing the People of the State of California, ... the general public whose interests must be served to justify an award of fees are the same citizens and residents of the state whose benefits disqualify their sovereign from an award.” Thus, in the view of the county defendants, attorney fees are never recoverable under Code of Civil Procedure section 1021.5 “in any public enforcement action brought by the Attorney General in his capacity as guardian of the public interest”; or, at the very least, such fees were not recoverable in this particular case because “the burden of this litigation and the benefit to the public interest do not transcend plaintiffs’ own interests that are served here.”

This argument has merit. Historically, the financial burden criterion of Code of Civil Procedure section 1021.5 has served to limit fee awards under the statute to persons who pursue public interest litigation at a cost to themselves that is out of proportion to any personal interests they might have in the outcome of the matter. The private attorney general doctrine embodied in Code of Civil Procedure section 1021.5 “rests on the recognition that in our complex society, citizens in great numbers frequently have interests in common that, while of enormous significance to the society as a whole, do not involve the fortunes of a single individual to the extent necessary to encourage their vindication by private recourse to the courts. Although there are offices and institutions within the executive branch of the government whose function is to represent the general public in such matters and to ensure proper enforcement (e.g., the Attorney General’s office), those offices and institutions are not always able adequately to carry the burden of enforcement, rendering private action socially useful. The issues involved in such litigation, are often extremely complex and their presentation time-consuming and costly. The award of substantial attorney’s fees to public interest litigants and their attorneys (whether private attorneys acting pro bono publico or members of ‘public interest’ law firms) who are successful in such cases encourages the representation of deserving interests and worthy causes.” (Save El Toro Assn. v. Days (1979) 98 Cal.App.3d 544, 552 [ 159 Cal.Rptr. 577 ]; accord, Serrano v. Priest, supra, 20 Cal.3d at p. 44 .)

Thus, an award of attorney fees under Code of Civil Procedure section 1021.5 has always served “as a ‘bounty’ for pursuing public interest litigation, not a reward for litigants motivated by their own interests who- coincidentally serve the public.” (California Licensed Foresters Assn. v. State Bd. of Forestry, supra, 30 Cal.App.4th at p. 570 .)

With the purpose of the financial burden criterion properly understood, a fundamental anomaly arises in applying that criterion in an action (like this one) brought by the Attorney General on behalf of the People of the State of California. The Attorney General needs no encouragement to pursue litigation that is in the general interest of the state’s population because, put simply, that is his or her job. As we have noted, under section 12511 (with certain exceptions not applicable here), “[t]he Attorney General has charge, as attorney, of all legal matters in which the State is interested . . . .” And as the court recognized in Save El Toro Assn. v. Days, supra, 98 Cal.App.3d at page 552 , the Attorney General’s office is one of the “offices and institutions within the executive branch of the government whose function is to represent the general public in such matters and to ensure proper enforcement.” (Accord, Serrano v. Priest, supra, 20 Cal.3d at p. 44 .) Rewarding the Attorney General with attorney fees for simply doing his or her job would essentially write the financial burden criterion, as it has historically been understood, out of the statute—at least where the Attorney General is concerned.

Significantly, the People do not dispute the historical understanding of the financial burden criterion, nor do they claim to be “exempt from meeting all the requirements of eligibility for fees under Section 1021.5 to make a successful claim.” Instead, they argue the financial burden criterion must be applied differently to public entities than it has historically been applied to private entities. They suggest that where a public entity is concerned, the financial burden criterion should involve an analysis of whether “the quantifiable costs of conducting the lawsuit” are out of proportion to any “pecuniary” interest the public entity has in the matter. Under this approach, “where the primary—or only—goal of a public entity’s suit is the enforcement or protection of a public policy interest, and the pecuniary return from the suit is little or nothing, the burden/benefit analysis should ordinarily come out in favor of eligibility for an award of fees.”

The People contend this different approach is required because otherwise “the rationale employed by the County and the trial court necessarily and always will preclude a public entity from recovering attorney fees under Section 1021.5, contrary to the express language in the 1993 amendment.” We simply cannot agree. There is nothing inherently inconsistent between the financiad burden criterion as it has been applied historically and an award of fees to a public entity, at least where that public entity is something less than the state itself.

Two cases will illustrate our point. As we have explained, in County of Inyo v. City of Los Angeles, supra, 78 Cal.App.3d at page 82 , this court addressed a motion for attorney fees by Inyo County in an action against the City of Los Angeles. (Id. at pp. 84-85.) In denying the county a fee award under the private attorney general doctrine, this court explained that the statewide benefits the county claimed it had achieved by its litigation were not “disproportionately important and valuable in comparison to [the county’s] own.” (County of Inyo, at p. 90.) The court went on to explain as follows: “A county is a political subdivision of the state which provides state and local governmental services for its inhabitants. [Citation.] Inyo County went to court as champion of local environmental values, which it sought to preserve for the benefit of its present and future inhabitants. This action is not a ‘public interest’ lawsuit in the sense that it is waged for values other than the [county’s]. The litigation is self-serving. The victory won by the county in 1977 bulked large enough to warrant the cost of winning it. The necessity for enforcement by Inyo County did not place on it ‘a burden out of proportion to [its] individual stake in the matter.’ ” (Ibid.)

In the second case—City of Hawaiian Gardens v. City of Long Beach (1998) 61 Cal.App.4th 1100 [ 72 Cal.Rptr.2d 134 ]—Hawaiian Gardens successfully sued to prevent Long Beach from closing a road at the border between the two cities, which would have diverted traffic onto residential streets in Hawaiian Gardens arid increased the likelihood of accidents on a street bordering a park and an elementary school; (Id. at pp. 1100, 1106.) On appeal from the trial court’s denial of attorney fees under Code of Civil Procedure section 1021.5, the appellate court affirmed, noting: “The record establishes that Hawaiian Gardens and its citizens received a substantial benefit when the proposed closure of Pioneer Boulevard was blocked. We agree with the trial court that while the judgment was of regional benefit, there is no showing that the burden of the litigation transcended Hawaiian Gardens’ interest in the controversy. The case was tried primarily on the administrative record compiled by Long Beach before the resolution was adopted. There was a brief hearing before the trial court.” ( 61 Cal.App.4th at p. 1113 .)

Although no fee award was made in either of these cases, the cases show that the traditional financial burden criterion in, Code of Civil Procedure section 1021.5 can easily be applied to public entities that are political subdivisions of the state—that is, something less than the state as a whole. In such a case, the pertinent question is whether the public entity deserves a reward for pursuing litigation that was in the interest , of a greater spectrum of the public than its own constituents. Although neither Inyo County nor the City of Hawaiian Gardens received a fee award, it was at least conceivable that they could have, if the cost of their litigation had transcended the benefits they secured for themselves, transforming it into true “public interest” litigation.

Thus, applying the traditional financial burden criterion to public entity litigants will not always preclude a fee award under Code of. Civil Procedure section 1021.5, except when the public entity litigant is the state itself, acting through the Attorney General. Such a case will always be self-serving, in that the People will always be pursuing their own interests through their chief attorney, whose very raison d’étre. is to enforce the laws of the state and serve the public interests of the state’s population as a whole. As the trial court recognized here, even if “the public interest of the state at large has been well-served by this litigation,” “serving the interests of the state at large is the public responsibility of’ the Attorney General. To reward the Attorney General with attorney fees for pursuing litigation it is his or her duty to pursue would stand the private attorney general doctrine on its head.

For the foregoing reasons, we conclude the People are not entitled to recover attorney fees in this action under Code of Civil Procedure section 1021.5, and the trial court erred in concluding otherwise.

DISPOSITION

The judgment is affirmed. The “Order Re Costs And Attorney Fees” is reversed to the extent it awarded attorney fees to the People pursuant to Code of Civil Procedure section 1021.5, but is otherwise affirmed. The parties will bear their own costs on appeal. (Cal. Rules of Court, rule 8.276(a)(4).)

Davis, Acting P. J., and Nicholson, J., concurred.

On April 11, 2007, the opinion was modified to read as printed above.

APPENDIX

AJb'WSMiUX

SECTIONS 10 & 15 1903 to 1904

[[Image here]]

SECTIONS 10 & 15 1904 to 1944

[[Image here]]

SECTION 34 1907 to 1910

[[Image here]]

SECTION 34 1910 to 1951

[[Image here]]

SECTIONS 32 & 5 1885 to 1936

[[Image here]]

SECTIONS 32 & 5 1944 to 1948

[[Image here]]

All further statutory references are to the Government Code unless otherwise indicated. These are the County of Tehama, the Tehama County Board of Supervisors, and Tehama County Planning Director George Robson. Because the second cause of action was the only cause of action in the third amended complaint directed against KAKE, it would have sufficed if KAKE had filed a motion for summary judgment only. The People later sought additional fees under Code of Civil Procedure section 1021.8, but the trial court denied that request on the ground it was “based on an unconstitutional purported amendment to” the statute. The People do not contest that ruling on appeal. Both notices of appeal purport to appeal from a postjudgment order entered on May 11, 2005. That was the date the notice of entry of order was filed; the order itself was filed on May 3. Because no one has complained and it is apparent what order the parties intended to appeal, we will liberally, construe the notices of appeal and treat them as referring to the “Order Re Costs And Attorney Fees” entered May 3, 2005. (See Cal. Rules of Court, rule. 8.100(a)(2).) KAKE also contended that to the extent the second cause of action was premised on an alleged violation of section 51257 of the Williamson Act, the cause of action was time-barred by the statute of limitations in subdivision (c)(1) of section 65009. KAKE does not renew this argument on appeal. Accordingly, we address only the limitations issue under the Subdivision Map Act. “When California became a state, federal surveyors divided it into a rectangular grid centered around one of three base and meridian landmarks. . . . Each of the townships in the grid is formed by the intersection of township and range lines and is identified by its position relative to [one of] th[es]e central point[s].... A township is approximately 6 miles square and consists of 36 sections, each about a mile square and containing 640 acres.” (Pleasant Valley Canal Co. v. Borror (1998) 61 Cal.App.4th 742, 755, fn. 7 [ 72 Cal.Rptr.2d 1 ].) No party agrees with the trial court’s conclusion that there were either two or four parcels in Burr Valley Estates immediately before the lot line adjustment. KAKE actually contends the parties agree there was one preexisting parcel in section 32 and disagree only on the number of preexisting parcels in section 5. This misconception arises from an error in the People’s papers. In their memorandum in support of their cross-motion for summary adjudication, the People asserted at two points that there was one parcel in section 32 and one parcel partly in section 5 and partly in section 32. At another point, they asserted there was “just one parcel in section 5 and one parcel, partly in section 5 and partly in section 32, prior to the so-called Lot Line Adjustment.” The People repeat this inconsistency in their appellate brief. Nonetheless, it is apparent the People rely on Blodger’s analysis with regard to sections 5 and 32, and Blodger concluded there were two parcels in these sections—one lying entirely in section 5 and another lying partly in section 5 and partly in section 32. Thus, notwithstanding KAKE’s assertion that the disagreement pertains only to section 5, we will analyze both section 5 and section 32. A patent is “ ‘[a] grant made by a government that confers on an individual fee-simple title to public lands.’ ” (Kellogg v. Garcia (2002) 102 Cal.App.4th 796, 800, fn. 1 [ 125 Cal.Rptr.2d 817 ].) Part of section 10—the north half of the northeast quarter and the northeast quarter of the northwest quarter—-was never part of the property now making up Burr Valley Estates, and we do not concern ourselves with that part of the section. “A metes and bounds description gives the boundary lines of- the property with their terminal points and angles. Essentially, it is any description other than by reference to particular lots and parcels on a recorded subdivision map.” ( 74 Ops.Cal.Atty.Gen. 149 , 150, fn. 2 (1991).) Specifically, the deed conveyed “all that certain lot and parcel of land, situate, lying and being in the County of Tehama, State of California, and bounded and particularly described as follows, to wit: [ffl Beginning at a point in the center of the Ridge Road from which point the comer to sections 10-11-14 and 15, in Township 26 N. R. 5 W. M. D. M., bears South 17.45 chains, thence following along the center of said Ridge Road by the following courses and distances: North 64° 30’ West 5.40 chains, South 77° West 3.57 chains, South 62° West 3.72 chains, North 73° West 3.17 chains, North 57° West 6.15 chains, North 71° 30’ West 7.48 chains, South 51° West 8.40 chains, South 50° 30’ West 900 chains to the center line of said Section 10, thence North on said center line 46.57 chains, more or less, to the North West comer of the South Half of the North East Quarter of Said Section 10; thence East 40.00 chains, more or less, to the line between Sections 10 & 11, thence South on said Section line 42.55 chains, more or less, to the place of beginning, containing 160 7/10 acres, more or less.” Indeed, KAKE asserts elsewhere that “land owners in the late nineteenth and early twentieth centuries could never have predicted” “the State’s increasing control over, and restriction of, development.” Given this assertion, it is difficult to imagine why the parties to the 1904 transaction would have been concerned about maintaining existing patent parcel boundaries as a means of protecting themselves against future regulation they “could never have predicted.” For the same reason, we are not concerned with the third deed in the land exchange, which involved Tinkham’s and Bell’s conveyance of the remainder of their property in section 10 south of Ridge Road to Bell. One of the federal patents included land outside of section 34, but the parcel created by that patent was later divided when the portion of the parcel within section 34 was conveyed to another owner, and only the portion in section 34 is part of the property at issue here. Specifically, the deed conveyed “all that certain lot and parcel of land, situate, lying and being in the County of Tehama, State of California, and bounded and particularly described as follows, to wit: [][] Beginning at the south-east comer of Section Thirty-four (34) in Township Twenty-seven (27) North of Range Five (5) West M. D. B. & M. Thence north 35.22 chains to the center of the County Road:—thence following along the center of the County Road, North 74 1/2° West 4.48 chains: thence South 74° West 5.00 chains:—Thence North 88 3/4° West 6.12 chains: thence North 79 1/2° West 5.20 chains:—Thence North 45° West 4.00 chains. Thence North 82 1/4° west 2.56 chains:—thence south 72° West 1.97 chains:—thence south 38.69 chains to the south line of said section Thirty-four (34):—thence east along said line 27.54 chains to the place of beginning, and containing 100 acres.” Specifically, the deed conveyed “all that certain lot and parcel of land, situate, lying and being in the County of Tehama, State of California, and bounded and particularly described as follows, to wit: [f] Beginning at a point on the South line of Section 34, in Township 27 North, of Range 5 West M. D. M., said point being 27.54 Chains West of the South east Comer of said Section 34, thence North 38.69 Chains to the Center of the County Road, Known as the ‘Johnson Road’, thence following along the Center of said Road South 71° West 4.35 Chains thence North 6854° West 2.77 Chains, thence North 4954° West 4.64 Chains, thence North 74° West 2.23 Chains, thence North 88° West 8.25 Chains, thence North 82 3/4° West 4.00 Chains thence South 43.10 Chains to the South line of said Section 34, thence North 89 1/4° East 24.57 Chains to the place of beginning, and Containing 100 acres, more or less.” Because the trial court correctly concluded the lot line adjustment had to comply with the Subdivision Map Act, we need not address the People’s alternate argument that, even if the lot line adjustment was exempt from the Subdivision Map Act, findings were required under section 51257 of the Williamson Act. It is unclear from the statutory language whether Code of Civil Procedure section 102115 actually limits a public entity to recovering attorney fees from a party on the bther side of the litigation that is also a public entity. The statute specifies that it “applies to allowances .. in favor of, public entities ... [if] one or more successful parties and one or more opposing parties are public entities . . . .” We need not decide that issue here, however, because the People did not seek fees from KAKE, but only from the county defendants. In bringing the action, the Attorney General claimed he was acting within his authority under section 12511, which, with certain exceptions not applicable here, provides that “[t]he Attorney General has charge, as attorney, of all legal matters in which the State is interested . . . .” According to the complaint, the Secretary of the State Resources Agency and the Director of the State Department of Conservation were acting pursuant to section 16147, which provides that “[t]he Secretary of the Resources Agency may request the Attorney General to bring any action in court necessary to enforce any enforceable restriction as defined in Section 422 of the Revenue and Taxation Code, upon land for which the secretary has certified payment of state funds to the local governing body during the current or any preceding fiscal year.” In addition, the Attorney General, the secretary, and the director all claimed to be acting pursuant to section 66499.33, which permits “any aggrieved . . . public agency” to “file a suit in the superior court of the county in which any real property attempted to be subdivided or sold, leased, or financed in violation of [the Subdivision Map Act] ... is located, to restrain or enjoin any attempted or proposed subdivision or sale, lease, or financing in violation of [the Subdivision Map Act].” The County of Inyo case, which obviously predated by many years the 1993 amendment to Code of Civil Procedure section 1021.5 that first allowed public entities to seek attorney fees under that statute, was decided under the common law private attorney general doctrine recognized in Serrano v. Priest, supra, 20 Cal.3d at page 25 before Code of Civil Procedure section 1021.5 was enacted.