dissenting.
Although the only chemical analysis conducted in this case showed a blood alcohol content (BAC) of .064 percent, the majority opinion goes to great lengths to avoid the plain meaning of the term “chemical analysis” in ORS 813.010(l)(a) in order to permit the state to use retrograde extrapolation to prove that defendant drove while his BAC was over .08 percent. Because I conclude that the legislature enacted ORS 813.010(l)(a) as a bright-line rule intended to include only those persons who actually register a BAC of .08 percent or greater on a chemical test, I respectfully dissent.'
When prosecuting a DUII, the state ordinarily may proceed under either paragraph (a) or (b) of ORS 813.010(1) (or both, in the alternative) — that is, the state may convict the defendant by proving either that the defendant drove with a BAC of .08 percent or more “as shown by chemical analysis,” or that the defendant was “under the influence of intoxicating liquor.” However, because of the limited procedural posture of this case, we consider the use of retrograde extrapolation only as it applies to paragraph (a) of ORS 813.010(1). Nothing in this opinion affects the use of that evidence under paragraph (b), as evidence of actual impairment at the time of driving. As a practical matter, then, the only case in which the state would require that evidence to secure a conviction is when a person both has a BAC lower than .08 percent when tested and has exhibited no outward signs of impairment sufficient to convict under paragraph (b). The majority holds that, in that situation, the statute permits the state to convict on the basis of expert testimony extrapolating from the results of a chemical analysis by factoring in additional, nonchemical evidence regarding mathematical formulas, average dissipation rates,1 and the time defendant last consumed alcohol and *16last drove, in direct contradiction to the statute’s directive that BAC may be proved only “by chemical analysis.” In my view, the more logical explanation is that the legislature simply intended for a person in that position to be acquitted. As I discuss in more detail below, both the structure of the statute and the plain meaning of its text requiring the requisite minimum BAC to be “shown by chemical analysis” support that conclusion.
As I interpret it, the structure of the statute indicates that there is only one crime — driving under the influence of intoxicants — -for which the legislature has provided two methods of proof. The first method of proof is to secure a chemical analysis showing a BAC above .08 percent. The second method of proof is to prove beyond a reasonable doubt that the defendant’s driving was actually impaired by the consumption of alcohol. By contrast, the majority’s interpretation has created an additional crime of driving with a BAC above .08 percent, the elements of which the state may establish by whatever means it pleases. 2 After today, the state need not be bothered to try to prove that any defendant’s driving was actually impaired (the actual wrongful conduct intended to be punished), unless the defendant’s BAC, when tested, was so low that by no possible “range” of extrapolation could the defendant’s BAC ever have exceeded .08 percent.
Because the first method of proof provided in paragraph (a) is a type of evidentiary presumption — a means by which the state is permitted to prove the commission of the offense without proving the actual underlying wrongful *17conduct — it would make sense for the legislature to limit the means of establishing that presumption to a method erring slightly on the side of under inclusivity. By limiting the use of paragraph (a) to those defendants who manifest a .08 percent BAC or higher at the time of the test, paragraph (a) of the statute is under inclusive as to those persons who may have been at or slightly over .08 percent BAC at the time of driving, but whose BAC has dissipated to below .08 percent by the time of the test. However, the state may still opt to prosecute those defendants under paragraph (b) by proving that their driving was actually impaired. Indeed, the state may still introduce its retrograde extrapolation evidence to demonstrate how much higher the defendant’s BAC likely was at the time of the crime, in conjunction with any other evidence tending to prove impairment. That interpretation of the statute, in which paragraph (a) may be used to establish a presumption of impairment in the most clear-cut cases, while paragraph (b) is used to ensure that in borderline cases wrongful conduct is established beyond a reasonable doubt, reliably encompasses the harm sought to be proscribed while adequately balancing the interests of both defendants and the state.
On the other hand, if, as the majority holds, retrograde extrapolation standing alone is sufficient to establish a violation of the statute, the statute risks over inclusivity, because there is no limit to how far the state may extrapolate to secure a conviction. The majority’s interpretation would permit the state to prove through retrograde extrapolation not only that a person with a .07 percent BAC was over .08 percent one hour earlier, but also that a person with a .01 percent BAC was over .08 percent eight hours earlier. The scientific principles supporting retrograde extrapolation have not advanced to such a state today — and certainly had not in 1975 when the legislature first enacted this statute — to permit such a conviction beyond a reasonable doubt. Indeed, in 1975, a first offense for driving under the influence of intoxicants was a mere traffic infraction, punishable by fine only, not to exceed $1,000. See former ORS 487.540(2) (1975), repealed by Or Laws 1983, ch 338, § 978 (providing that “[d]riving while under the influence of intoxicants is a Class A traffic *18infraction”); former ORS 484.360(1), (2) (1975), repealed by Or Laws 1983, ch 338, § 978 (providing that “the penalty for committing a traffic infraction shall be a fine only” which, for Class A traffic infractions, may not exceed $1,000). It can hardly be contended that in that context the 1975 legislature could have envisioned a battle of competing scientific expert witnesses at a full-blown criminal trial testifying with regard to science that, if it existed at all, was in its very nascence. That the legislature intended to create a simple, bright-line evidentiary presumption is more plausible in that context.
The plain meaning of the text of the statute also supports my conclusion that the legislature had no intention of including retrograde extrapolation as a “chemical analysis” under paragraph (a). The legislature required that a BAC of .08 percent or more must be “shown by chemical analysis of the breath or blood of the person” in order to support a determination that a person has committed the offense of driving under the influence under ORS 813.010(l)(a). The legislature did not provide that a BAC of less than .08 percent, as shown by chemical analysis, could somehow be rehabilitated or explained away by expert testimony or any other circumstantial evidence under paragraph (a). The terms the legislature used in paragraph (a) plainly require that a chemical analysis made under ORS 813.100, ORS 813.140, or ORS 813.150 must itself show that a person has a BAC of .08 percent or greater. That reading of the text is well supported by the applicable definitions of the terms the legislature used in the statute and by the general rules of grammatical construction that inform our understanding of statutory terms.
The statutory phrase, “as shown by chemical analysis of the breath or blood of the person made under ORS 813.100, 813.140 or 813.150,” when broken down into its significant constituent parts, provides the following construct. “Shown” is the past participle of the verb “show,” and the synonyms for the general term “show” are as follows:
“syn EVINCE, MANIFEST, EVIDENCE, DEMONSTRATE: in this series SHOW is a general term, usu. interchangeable with any of the others, for indicating, revealing, displaying!.]”
*19Webster’s Third New Int’l Dictionary 2105 (unabridged ed 2002). “Chemical” means: “3 : having reference to or relating to the science of chemistry!.]” Id. at 383. “Analysis,” in the context of chemistry, means:
“4 a : the separation of compound substances into their constituents by chemical processes b : the determination, which may or may not involve actual separation, of one or more ingredients of a substance either as to kind or amount; also : the tabulated result of such a determination [.]”
Id. at 77. Thus, the phrase “as shown by chemical analysis of the breath or blood of the person” most directly and most reasonably is read to mean the result demonstrated by the chemical test itself.
That point is made even more evident by the legislature’s explicit articulation of the three specific ways in which a person’s BAC may be “shown” under paragraph (a): by a chemical analysis “made under ORS 813.100, 813.140 or 813.150.” Those statutory provisions provide for the administration of chemical tests of the defendant’s breath, blood, or urine — without mention of any kind of extrapolation or extrinsic interpretation of the results of those tests. In limiting the chemical analyses that can be used to those “made” under those statutes, the legislature specified a manner of proof — a specific type of test with a specific result. It is the office of this court to ascertain and declare what is contained within the substance of the statute — not to insert what may have been omitted or to second-guess the policy choices made by the legislature. See ORS 174.010 (providing that “[i]n the construction of a statute, the office of the judge is simply to ascertain and declare what is, in terms or in substance, contained therein, not to insert what has been omitted.”). Whether or not we think that modern retrograde extrapolation analysis may be a reliable and convenient manner of proving a defendant’s BAC at the time of driving, it is not for us to add a fourth method to the three the legislature saw fit to permit. Consequently, I determine that we should take the terms the legislature enacted in ORS 813.010(l)(a) at face value and adhere to the policy choice that the legislature made in setting out the specific manner for proving a violation of the statute.
*20The legislative history of the statute also supports that interpretation. As the majority notes, the legislative history of ORS 813.010(l)(a) provides little direct enlightenment regarding whether the legislature ever actually considered the problem of dissipation in the interim between the time of driving and the time of a chemical analysis. 352 Or at 14 n 10. But what little legislative history we do have on the subject is at least consistent with, if not supportive of, a more restrictive reading of ORS 813.010(l)(a) than the majority allows.
The text “as shown by chemical analysis” was initially enacted in former ORS 487.540(l)(a) (1975), repealed by Or Laws 1983, ch 338, § 978, as part of a comprehensive overhaul of Oregon’s motor vehicle laws, which included changes to Oregon’s DUII laws.3 Prior to enacting that statute, the Committee on the Judiciary, which prepared the Motor Vehicle Code and the text of former ORS 487.540(l)(a), discussed the evidentiary problems that alcohol absorption and elimination could create in DUII prosecutions. According to the committee minutes, then-Solicitor General W. Michael Gillette testified before the committee regarding the problem with time lapses in chemical analyses in the following terms:
“[Gillette’s] concern was that the trier of fact had been told that some use could be made of a blood alcohol content of .08 percent and lower as indirect evidence but he was not told what to do with evidence of .08 percent or higher. In addition to instructions as to what to do when the evidence was less than .08, his proposal would give the courts direction as to what to do when the analysis showed more than .08. He pointed out that the blood test was not contemporaneous with the driving; the test results were established at a later time. His answer to this problem as set out in his proposed revision *** was that a person who was shown to have .08 percent or higher at the time of the blood test was at least then under the influence, and from there it could be inferred *21backward that he was under the influence at the earlier time, even though the court could not be certain that his blood alcohol content at the time of driving was .08.”
Minutes, Committee on Judiciary, Sept 24,1974,7 (emphasis added). That statement highlights that the proposed answer to the evidentiary problems that alcohol absorption and elimination could create was to provide that, once a person is shown to have a BAC of .08 percent or higher at the time of the test, it can be inferred backward that the person must have had a BAC no lower than .08 percent at the time of driving — even though the court could not be certain of the defendant’s precise BAC at the time of driving.4 That theory is consistent with the language of the statute actually enacted by the legislature, requiring BAC to be proved under paragraph (a) “by chemical analysis,”5 and also comports with the additional provisions enacted by the legislature as former ORS 487.545(1) and (2) (1975).6 Former ORS *22487.545(1) specifically provided that the appropriate use of a BAC lower than the legal limit (then .10 percent), “as shown by chemical analysis,” was as “indirect evidence that may be used to determine whether or not [defendant] was then under the influence of intoxicants.” By contrast, former ORS 487.545(2) provided that a BAC over the legal limit “constitutes being under the influence of intoxicating liquor.”
Consistent with that legislative history, all of the statutory provisions can be given full effect by determining that the requisite proof for a prosecution under ORS 813.010(l)(a) is a chemical breath or blood test showing a BAC of .08 percent or higher and determining that a chemical test showing a BAC lower than .08 percent constitutes only indirect evidence that the person is under the influence of intoxicants for prosecutions under ORS 813.010(l)(b) or (c). Those determinations, of course, lead ineluctably to the conclusion that a chemical test showing a BAC lower than .08 percent cannot establish that a person drove with a BAC of .08 percent or higher, and any testimony about the import of such a test result (based upon retrograde extrapolation or any other theory) is not relevant for purposes of determining whether the person violated ORS 813.010(l)(a).
Furthermore, this determination also comports with the prior decisions of this court. In State v. O’Key, 321 Or 285, 308, 899 P2d 663 (1995), this court stated that “the offense of DUII with a .08 percent or more BAC may be proved only by a ‘chemical analysis of the breath or blood of the person made under ORS 813.100, 813.140 or 813.150.’” (Emphasis in original.) The majority correctly notes that the precise issue addressed by this court in O’Key was limited to whether the horizontal gaze nystagmus (HGN) test was admissible under ORS 813.010(l)(a) to prove that the defendant drove with a BAC of .08 percent or more, but I think it significant that this court described the universe of available proof under ORS 813.010(l)(a) as limited to a chemical analysis of the breath or blood. The majority opinion distinguishes in this case between an HGN *23test, which is clearly not a chemical test, and testimony concerning retrograde extrapolation, which the majority concludes is derived from a chemical test. 352 Or at 11-12. But in doing so, the majority opinion acknowledges, as it must, that testimony explaining retrograde exploitation analysis is not a chemical test. Id. And that is where we draw different lines. While the majority opinion concludes that testimony concerning retrograde extrapolation is admissible to prove a violation of ORS 813.010(l)(a), despite the fact that the majority agrees that testimony about retrograde extrapolation is not a chemical test that shows the BAC of the defendant, I conclude, in harmony with this court’s express statement in O’Key, that only a chemical test that shows that the BAC of the defendant was .08 percent or more at the time of the test is admissible to prove such a violation.
The majority also relies in part on this court’s decision in State v. Clark, 286 Or 33, 593 P2d 123 (1979), for support. 352 Or at 10-11. The only issue directly presented in Clark, however, was defendant’s right to attack the validity and accuracy of the chemical test result that showed that defendant had a BAC of .10 percent or more under former ORS 487.540. The narrow issue decided by this court in Clark was that,
“because the ‘chemical analysis’ of the blood test of this defendant by the use of a Breathalyzer could ‘produce invalid results through human error or chemical or mechanical defect,’ the defendant was entitled to attempt to impeach the accuracy of the testimony of the expert witness relating to the results of that test by any competent evidence.”
286 Or at 43. Here, of course, the testimony that the state wished to present concerning retrograde extrapolation was not in any way intended to impeach the accuracy of the chemical test result; indeed, the testimony necessarily relied upon the accuracy of the chemical test result that showed that defendant’s BAC was .064 percent (well below the statutory standard of .08 percent). Nothing expressed by this court in Clark addresses the issue presented in this case — viz., whether evidence that is unrelated to the accuracy of the chemical test result is admissible in a prosecution under ORS 813.010(l)(a).
*24Finally, the majority also notes this court’s decision in State v. Parker, 317 Or 225, 855 P2d 636 (1993). 352 Or at 9-10. In Parker, however, there was no distinction drawn between a prosecution pursued under ORS 813.010(l)(a) and prosecutions pursued under ORS 813.010(l)(b) and (c). See Parker, 317 Or at 227 (noting that defendant was charged with driving under the influence under ORS 813.010 generally). The court then merely noted in passing, in addressing the propriety of a denial of a request for a continuance, that in a prosecution brought under ORS 813.010 the state could rely upon the common knowledge that blood alcohol dissipates over time, and that defendant could have prepared to address that issue whether or not the state gave timely notice of its intent to present expert testimony on the matter. Id. at 232 n 9. In addition, while this court did also refer to the fact that there was expert testimony presented to establish that defendant’s BAC was greater than .08 percent at the time of the accident, based upon a breath test result showing a BAC of .07 percent more than five hours later, that discussion is of no moment here, because the admissibility of that evidence in a prosecution pursued solely under ORS 813.010(l)(a) was not presented or decided in Parker. Consequently, this court’s decision in Parker is also inapposite to the issue presented in this case.
As discussed above, I conclude that the text and legislative history establish that the legislature intended prosecutions under ORS 813.010(l)(a) to rely only on chemical test results showing that the defendant had a BAC of .08 percent or more at the time of the test. I also conclude that that interpretation of ORS 813.010(l)(a) is consistent with, and not precluded by, this court’s prior decisions. Consequently, I would hold that the trial court properly excluded the retrograde extrapolation testimony proffered by the state. I respectfully dissent, therefore, from the contrary determination reached by the majority.
Durham and Walters, JJ., join this opinion.The mathematical formulas used in the kind of retrograde extrapolation performed in this case were not specifically tailored to the defendant, but were derived from studies of average alcohol dissipation rates across a broad population. As the state’s expert admitted, those rates can vary widely from person to person, preventing the analysis from determining what defendant’s specific BAC was at the time of driving. Rather, the tests can return only a “range” of possible results, which, as the state’s expert testified, will be accurate for only 98 out of 100 people. 352 Or at 5 n 3. In my view, using those averages and calculations that *16are not specific to defendant’s particular physiology and circumstances directly contravenes the statutory text that the subject of the “chemical analysis” must be the breath or blood of the defendant. ORS 813.010(l)(a).
It is no defense to the majority’s position that the retrograde extrapolation analysis is “based” on a chemical analysis. There is no way to arrive at any known percentage of blood alcohol content — as the statute requires — without some reference to a chemical analysis. The inclusion of the additional requirement in the statute that the .08 percent must be “shown by chemical analysis of the breath or blood of the person made under ORS 813.100, 813.140 or 813.150” is better understood not as redundant to that obvious truth but rather as limiting the universe of methods by which the state is permitted to prove the defendant’s precise blood alcohol content. That limitation necessarily implies that there are other ways to prove that defendant’s blood contained .08 percent alcohol at the time of driving; however, only those ways expressly listed are admissible to satisfy the evidentiary presumption provided by paragraph (a).
Although former ORS 487.540 was repealed by Or Laws 1983, ch 338, § 978, the legislature’s intent with regard to that provision presumably applies with equal force to ORS 813.010(l)(a), which was enacted by a different section of that same chapter and contained substantially the same text. The legislature explicitly provided that “[i]t is not the purpose or intent of the Oregon Legislative Assembly to change the law by enacting the revision of the Oregon Vehicle Code contained in chapter 338, Oregon Laws 1983.” Or Laws 1983, ch 338, § 3.
Solicitor General Gillette’s testimony referred to a BAC of .08 percent. The legislature initially enacted provisions in former ORS 487.545 (1975), repealed by Or Laws 1983, ch 338, § 978, and former ORS 487.540 (1975), that set the legal presumption of intoxication at .10 percent BAC. The legislature then later modified the standard to .08 percent BAC, as it is currently set out in ORS 813.300 and in ORS 813.010. Those changes in the BAC standard, however, do not affect the central issue in this case.
Former ORS 487.540(l)(a) (1975) provided:
“A person commits the offense of driving while under the influence of intoxicants if he drives a vehicle while:
“(a) He has .10 percent or more by weight of alcohol in his blood as shown by chemical analysis of his breath, blood, urine or saliva made under ORS 487.805 to 487.815 and 487.825 to 487.835[.]”
Former ORS 487.545(1) and (2) provided:
“(1) At the trial of any civil or criminal action, suit or proceeding arising out of the acts committed by a person driving a motor vehicle while under the influence of intoxicants, if the amount of alcohol in the person’s blood at the time alleged is less than .10 percent by weight of alcohol as shown by chemical analysis of the person’s breath, blood, urine or saliva, it is indirect evidence that may be used to determine whether or not he was then under the influence of intoxicants.
“(2) Not less than .10 percent by weight of alcohol in a person’s blood constitutes being under the influence of intoxicating liquor.”
ORS 813.300(1) and (2) now provide:
“(1) At the trial of any civil or criminal action, suit or proceeding arising out of the acts committed by a person driving a motor vehicle while under the influence of intoxicants, if the amount of alcohol in the person’s blood at the time alleged is less than 0.08 percent by weight of alcohol as shown by chemical analysis of the person’s breath or blood, it is indirect evidence *22that may be used with other evidence, if any, to determine whether or not the person was then under the influence of intoxicants.
“(2) Not less than 0.08 percent by weight of alcohol in a person’s blood constitutes being under the influence of intoxicating liquor.”