Clearly, the Ezaugi defendant recanted his false testimony only after being convinced that his perjury was no longer believable. Realizing this, the court of appeals considered Detective Ezaugi's recantation "not a demonstration of penitence to purge the torments of a guilty conscience, but a calculated effort to escape the dire consequences of admitted false swearing." [FN51] 

Taking these concerns into consideration, the Ezaugi court then limited the application of the recantation doctrine to the following circumstances: (1) when a perjurer corrects knowingly false testimony; (2) if it is done "promptly"; (3) if it is done "before the body conducting the inquiry"; (4) if it is done before the inquiry has been deceived or misled to the detriment of its investigation; (5) and, finally, if no reasonable likelihood exists that the perjurer has
learned his untruths have been or will be discovered. [FN52] 

The fourth and fifth elements were entirely new to New York (the fifth was identical to a concern expressed in Norris) and caused one dissenting justice to fear
the demise of the defense's utility in light of the majority's decision. [FN53] All other elements the Ezaugi court listed were inherited from Gillette. 

B. The Current Recantation Law in New York 

In 1965, the New York legislature codified the recantation doctrine based on the
Ezaugi decision in section 210:25 of the New York Penal Law. [FN54] In doing so,
the legislature made it an affirmative defense and adopted substantially the
language of the Model Penal Code's retraction statute. [FN55] Unlike New York's
common law recantation defense, New York Penal Law Section 210.25 called the
defense "retraction," rather than "recantation," and required that a witness retract
his false statement "in the course of the proceeding in which it was made," rather
than "promptly," as was held in Ezaugi. [FN56] 

The codified retraction defense, which to present has never been amended,
reads as follows: 

In any prosecution for perjury, it is an affirmative defense that the
defendant retracted his false statement in the course of the
proceeding in which it was made before such false statement
substantially affected the proceeding and before it became manifest
that its falsity was or would be exposed. [FN57]


1. New York Penal Law Section 215.25 is Unclear 

Probably the most ambiguous part of the statute is the term "in the course of the
proceeding." Does this mean during the course of an entire criminal trial? That is,
from grand jury to sentencing hearing, or just during the grand jury hearing? [FN58]
Or in a civil trial, from discovery until all post-trial appeals have been exhausted? 

Unfortunately, New York lawmakers left no legislative history - and the courts
have generated little precedent - to explain the duration of the locus poenitentiae,
[FN59] which, in the context of a recantation rule, refers to the time period in which
one may correct his misstatement in order to be pardoned from a perjury
prosecution. To complicate matters, the little precedent that exists is pre-1967
(before the statute was codified and enacted); thus, it interprets the common law
recantation defense, which uses the word "promptly," from Ezaugi, rather than "in
the course of the proceeding." Lastly, though the statue has been in effect and
good law for over thirty years, no court has rendered a published decision
commenting on the duration of the locus poenitentiae in reference to the
codification's language. 

Needless to say, it is unclear how applicable the pre-1967 recantation case law is
to today's statute. Keeping this in mind, this Article will now examine the few New
York decisions that define the locus poenitentiae. 

Early in the century, one court held the recantation defense to be viable when one
corrects false testimony "before the submission of the case." [FN60] Oddly
enough, the same court later barred the recantation defense for one who recanted a
four-month-old misstatement before the termination of the proceeding. [FN61] The
most recent case, decided in 1959 (which is the only one to comment on the
pre-codification Ezaugi standard), said a correction of testimony over two weeks
after a witness first lied to a grand jury did not automatically preclude the use of the
recantation defense. [FN62] 

From a plain reading of the statute, "in the course of the same proceeding" - if not
constituting the entire proceeding - is at least a longer period of time than "done
promptly." The few commentators that discuss this distinction concur: 

In place, however, of the Ezaugi requirement that the retraction be "done
promptly," S210.25 provides a defense if the retraction is made "in the course of the
proceeding." If there is a temporal difference between the making of the false
statement and a retraction thereof by the defendant, S210.25 recognizes that the
purposes of justice are equally well served if the retraction is something less than
"promptly" made, provided however, that when the false statement is retracted, it
has not substantially affected the proceeding and has not been or was not then
likely to be exposed. [FN63] 

As mentioned previously, to date no New York court has discussed this
distinction in the context of New York's retraction statute. But a few other
jurisdictions have either by statute or case law defined "procedure" within the
context of their recantation defense. [FN64] Perhaps one of these cases might be
persuasive to a New York court pondering this distinction. 

Ambiguity still remains, however, and liars deciding whether to correct a lie have
no clear-cut answer to whether their locus poenitentiae has expired. 

III. The Completed-Crime Rule: The Rejection of the Recantation Defense 

Jurisdictions that reject the recantation defense consider the act of making willful
and knowingly false statements to be criminally culpable behavior. "Deliberate
material falsification under oath constitutes the crime of perjury, and the crime is
complete when a witness's statement has once been made," said the Supreme Court
in United States v. Norris. [FN65] 

The key element to completing the crime of perjury is willfulness. Thus, a witness
whose conscience compels him to subsequently correct lies he has offered while
under oath is still a perjurer who deserves punishment, according to the
complete-crime rule. In fact, as the Norris court said, a witness's first willful
misstatement is considered culpable conduct from the instant it was uttered;
therefore, he cannot escape the penal consequences by invoking a defense. [FN66]

The Norris Court, an ardent supporter of the completed-crime rule, expressed its
distaste for the recantation defense: 

[The recantation defense] ignores the fact that the oath
administered to the witness calls on him freely to disclose the truth
in the first instance and not to put the court and the parties to the
disadvantage, hindrance, and delay of ultimately extracting the truth
by cross examination, by extraneous investigation(,) or other
collateral means. [FN67]


Proponents of the completed-crime rule, such as the Norris Court, feel its deterrent
value most effectively optimizes truthfulness of initial statements by deterring
fabrication in the first place, [FN68] in addition to punishing liars for culpable
behavior regardless of their subsequent corrections, if any. In other words,
retributive theory favors punishment for an offered lie, regardless of any retraction
made by the liar. [FN69] 

The completed-crime rule became known as the federal rule on recantation, as a
result of the Supreme Court's adoption of it in Norris. [FN70]Ironically, use of this
term today would be an anachronism since Congress substantially rejected the
completed-crime rule in the perjury section of its Organized Crime Control Act in
1970. [FN71] Despite Congress's adoption of the recantation rule, the majority of
states still adhere to the completed-crime rule. In fact, one completed-crime
jurisdiction expressly rejects the recantation rule by statute. [FN72] 

IV. A Survey of the Recantation Doctrine Nationwide 

All recantation defenses in the United States, whether they be court made or
statutory, [FN73] derive their basic elements from the defense as set forth in Ezaugi
and later codified by New York's legislature (which adopted substantially the
language of the Model Penal Code [FN74]). Although other jurisdictions' elements
are borrowed from New York, not all recantation defenses are the same. For
instance, some states use only a portion of New York's elements, while others use
them all; hence, they follow what has been known as the "New York rule." [FN75]
Of those elements that are borrowed, some jurisdictions use differing language.
Such language variations may be only subtle, yet they have an impact on the
defense's application. But all jurisdictions that subscribe to the defense are the
same in that all the elements used, in part or whole, come from New York. In other
words, states have not created new elements that are unique to their jurisdiction. 

There are three basic elements, which include (A) motive or mens rea, (B) locus
poenitentiae, and (C) effect on party and/or proceeding. 

A. Motive or Mens Rea Element 

Motive is the "cause or reason that moves the will and induces action." [FN76]
The first recantation defense, formulated by People v. Gillette, [FN77] made no
mention of a motive requirement for the recanter. Other subsequent decisions,
however, heavily criticize the Gillette court's failure to mention the motive element;
they insist that without one, the incentive to perjure oneself would actually
increase. [FN78] Today, most jurisdictions and the Model Penal Code have
expanded on Gillette and now look to the liar's mens rea to determine if he deserves
a defense. Keep in mind, however, that motive, in the context of this Article, does
not refer to the reason the witness originally lied on the stand. Rather, in the
context of the recantation doctrine, motive refers to the liar's reasons for recanting
his misstatements. In particular, a court would look to see if a recanter's motivation
for correcting his lies is to avoid prosecution by authorities who are aware or will
become aware of the lies. Although this goal is universal to most recantation-rule
jurisdictions, the language jurisdictions employ to achieve this goal is sometimes
different. Additionally, some jurisdictions - like the Gillette opinion - still disregard
motive entirely. [FN79] Below this Article will discuss the different language
jurisdictions use to determine whether the liar's motive for recanting entitles one the
shelter of a recantation defense. In addition, the Article shall look at recantation
statutes that disregard motive altogether and how motive, or a lack thereof, affects
application of the recantation defense. 

1. Objective-View-of-Motive Standard 

Most recantation defense language requires a false statement to be corrected 
"before it became manifest that the falsification (of one's prior statement) was or
would be exposed." [FN80] 

This type of language, semantically, does not look exclusively at a "pure motive" -
that is, the subjective reason the liar recants. The language aims to preclude the
recantation defense to a witness who fears he will soon be caught and hopes to
escape punishment by recanting, because the authorities have discovered or will
discover his untruths. 

The term "objective-view-of-motive" standard is more accurate because a sensible
interpretation of the language "it becomes manifest" does not require courts to look
only to the liar's mens rea. If the authorities merely have discovered the lie or will
discover the lie in the future, then the defense is barred. Admittedly, however, few
courts have discussed this distinction. In fact, some courts in interpreting "it
becomes manifest" have flatly stated the opposite; that is, that this language looks
only to the subjective mens rea of the liar. [FN81] Thus, this language, as
interpreted by these courts, is really a good-faith motive, which is discussed in the
following section. [FN82] 

Despite these interpretations, this Article contends that a plain reading of the
language instructs courts to look to the outside circumstances that exist at the time
of a liar's recantation to determine if he had the proper motivation to recant -
namely, whether authorities have discovered or will discover the lies before the liar
recants. [FN83] Thus, it is possible for a lying witness who is convinced of the
secrecy of his misstatements to repent and offer a retraction and still face a perjury
conviction. For instance, if, unknown to a now-recanting witness, authorities learn
through other means (or it is manifest that they will subsequently learn through
other means) that the witness lied when first testifying, the witness's recantation
defense is divested and he will face a perjury conviction. In order to fulfill this
element's language, the recantation must have been done "before it became
manifest that the falsity of one's prior statement was or would be exposed." [FN84]
This means manifest to the authorities, or, presumably, the liar, himself, believes it
has become manifest to the authorities. [FN85] Therefore, if the authorities discover
the lie, it is irrelevant what the recanter believes. In summary, under the
objective-view-of-motive standard, if either the authorities discover the lie, or the
liar subjectively believes the authorities have discovered or will discover the lie
(even if the authorities had not and will not discover the lie), [FN86] the recantation
defense is unavailable under the objective-view-of-motive. 

2. Good-Faith Motive Standard 

Jurisdictions that follow a good-faith standard provide that "a recantation must
take place before the discovery of the falsification became known to the witness,
himself." [FN87] Though the policy goal for this type of language is identical to
that of the objective-view-of-motive language discussed previously, [FN88]the
outcome of its application is not always the same. With this purely subjective
language, it is irrelevant whether the authorities ever learn of the lie. To invoke the
defense, it only matters that the liar himself has no knowledge that the authorities
have or will have learned of his lie before he retracts it. Thus, as long as a witness
believes his untruths are secret, he may invoke the defense. Needless to say, once
the authorities have made public their knowledge of the lies or commenced a
perjury prosecution against the liar, the discovery of the lie is known to the liar and
the defense is divested. 

This good-faith motive is better understood by contrasting it with the
objective-view-of-motive standard in a hypothetical: A witness testifies falsely, but
later decides to clear his conscience by offering the truth to the court. Unknown to
the witness and before his recantation, the district attorney obtains documents that
incontrovertibly prove the witness willfully lied while testifying. Using an
objective-view-of-motive, the witness must be convicted of perjury because it
became manifest that the falsification was or would be exposed. For an
objective-view-of-motive standard, the liar's beliefs are usually irrelevant. 

However, with a defense that requires a goodfaith motive, this element is satisfied
because the witness subjectively believes his false statements were secret at the
time of recantation. Therefore, under a good-faith motive requirement, the witness's
retraction excuses his former perjury. 

3. Motive Irrelevant 

A small number of jurisdictions have no motive requirement for their recantation
defenses. [FN89] For these jurisdictions, it is important only that the statement was
retracted - the reason why is purely irrelevant. As long as a liar recants, he may still
invoke the defense so long as all other requisite elements of the defense are
satisfied. 

Jurisdictions with this type of defense are few in number and have elicited
criticism from courts and commentators alike; [FN90] even some among those who
are ardent supporters of the recantation doctrine dislike a no-motive recantation
defense. [FN91] Clearly, those who subscribe to the complete- crime rule believe
that disregarding the motive of a recanter perverts justice by encouraging perjury.
[FN92] 

Nonetheless, the motive-irrelevant standard does have its advocates who reason
as Judge Desmond does below: 


[S]ince the recantation rule's purpose is not to reward or punish the
liar but to get the truth into the record, the perjurer's motive for
recanting has nothing to do with it at all. 
The high public purposes and policy behind the recantation rule
should constrain us to uphold and implement it, not destroy it by
limitations [such as a motive]. [FN93]


Even the American Law Institute, which advocates a motive standard in its Model
Penal Code, [FN94] concedes that there is "some possibility that the defense may
be unfairly denied if the courts apply too rigidly the requirement that recantation
precede exposure of the falsehood." [FN95] Nevertheless, the no-motive standard
is still regarded as unsound and followed by very few jurisdictions. 

B. Locus Poenitentiae or Time Period 

Locus poenitentiae, [FN96] Latin for "opportunity to repent," refers to the time
period in which one may recant false testimony and avoid a perjury prosecution.
Like all other recantation defense elements, a subtle difference in language affects
the defense's application significantly. The Ezaugi court first required a liar to
recant his statement "promptly" before it became manifest that the falsity was or
would be exposed and before the proceeding was prejudiced. [FN97] Following the
Model Penal Code's lead, [FN98] New York codified Ezaugi using the language "in
the course of the proceeding in which it was made," [FN99] instead of "promptly."
[FN100] The vast majority of jurisdictions use language that is largely synonymous
to the Model Penal Code. A smaller number of courts use language that is similar,
but more defined, such as during the "same continuous trial," [FN101] "before
completion of the testimony at the official proceeding," [FN102] and before the case
is "submitted to the ultimate trier of fact." [FN103] 

But for the Model Penal Code, and other statutes like it, what does "proceeding"
mean? As mentioned earlier, for over thirty years that New York Penal Law section
210.25 has been on the books, New York's judiciary has never commented on its
recantation defense statute. [FN104] Other jurisdictions have at least received
interpretations from their judiciaries: "Without question," a New Jersey court said,
"the term 'proceeding,' standing alone, is broad enough to cover each step or all
steps in a criminal action from commencement to final legislation." [FN105] 

Other courts have given the term "proceeding" a narrower reading than New
Jersey. [FN106] In addition, some state lawmakers have, themselves, expressly
defined what "proceeding" means by statute. [FN107] 

The definition of such words is of paramount importance. For example, if 
"proceeding" is construed narrowly, it could mean before the testifying witness
leaves the stand. In this instance, one who recants after leaving the stand may not
have caused harm to the parties or proceeding, and he may have even retracted
before it became manifest that his falsity has or would be discovered; yet because
his retraction was after the mandated locus poenitentiae (which in this instance is
the same "proceeding," interpreted to mean before he leaves the witness stand), his
recantation defense will fail. By contrast, jurisdictions that interpret "proceeding"
broadly, might award a recantation under these circumstances. 

Another important distinction is that jurisdictions employ one of the three
different forms of locus poenitentiae. Some use a fixed or independent time period
to determine whether a liar deserves a pardon. Others make the locus poenitentiae
contingent upon the motive and/or prejudice to a party or proceeding. Lastly, most
jurisdictions use a combination of both of these two. 

In the absence of accepted terminology to describe these variations, this Article
takes the liberty of coining terms for them: (1) "fixed locus poenitentiae," (2)
"contingent locus poenitentiae," and (3) "hybrid locus poenitentiae" respectively.
Each variation is discussed in turn. 

1. Fixed Locus Poenitentiae 

A small number of jurisdictions require a liar to recant before a finite period of time,
which is defined by the language comprising the defense. [FN108] Unlike the other
locus poenitentiae variations, this time period is not contingent upon any other
events. An example of such language can be found in Colorado's retraction statute:
"No person shall be convicted of perjury in the first degree if he retracted his false
statement in the course of the same proceeding in which it was made." [FN109]
Note that the only time-period requirement is that a witness retract his falsehood
"in the course of the same proceeding in which it was made." [FN110] 

2. Contingent Locus Poenitentiae 

After a careful reading of the language of some recantation statutes, it becomes
apparent that the locus poenitentiae is not always a finite period of time that is the
same under all circumstances. Instead, most recantation defenses have variable
time periods that are contingent upon another factor or factors. [FN111] For
instance, usually one may recant before the parties or proceeding are prejudiced,
and some defenses require a lie to be retracted before it becomes manifest that the
falsity of their testimony has or will be known. [FN112] Additionally, some
jurisdictions require a full recantation before both of these. The following
recantation statute is an example of a contingent locus poenitentiae: 

S X Recantation - Defense to Perjury 
It is a defense to perjury that a witness recant a knowingly false
statement before it becomes manifest that the falsity has been or will
be discovered or the lie has substantially prejudiced any party or the
proceeding. [FN113]

Accordingly, for a recanter to successfully invoke a section X defense, he must
recant before the lie has been or will be exposed or a party or the proceeding has
been prejudiced. 

Few courts, if any, have discussed the advantages or disadvantages of one locus
poenitentiae over another. One commentator, however, has considered the
distinction and, though not using this Article's terminology, expresses his
fondness for the contingent locus poenitentiae, while criticizing a fixed time period: 

[T]he immediacy with which testimony must be corrected in order
for the perjury to be excused should be construed to require
measurement not by an inflexible rule which perfunctorily rejects any
correction made after an arbitrarily determined period of time.
Instead, immediacy should be determined primarily by the measure
of inconvenience or prejudice which the witness's false testimony
has caused. [FN114]


Despite this commentator's endorsement, recantation defenses that employ a
purely contingent locus poenitentiae element are few, if any. 

3. Hybrid Locus Poenitentiae 

A hybrid locus poenitentiae has both a fixed period and a period that is contingent
upon other events. Of all the jurisdictions that advance the recantation rule, the
hybrid locus poenitentiae enjoys the most popularity. This is due, no doubt, to
New York's Ezaugi standard, [FN115] which the American Law Institute
promulgates in its Model Penal Code: 

Retraction. No person shall be convicted of an offense under this
(perjury) Section if he retracted the falsification in the course of the
proceeding in which it was made before it became manifest that the
falsification was or would be exposed and before the falsification
substantially affected the proceeding. [FN116]


Notice there is both a fixed time period in which one must recant and a variable
one that depends upon either the falsification being exposed or a party or the
proceeding being prejudiced. In the majority of recantation defenses, like the Model
Penal Code above, both of these preconditions are required before the opportunity
to recant expires. 

Thus, one may retract during the fixed period, but if the other event or events
upon which the locus poenitentiae is contingent occurs, the opportunity to recant
is divested. This is so even though the fixed time period, which is "in the course of
the same proceeding" for the Model Penal Code, may not have passed. Likewise, if
the fixed time period expires, one may not successfully recant if the events upon
which the contingency depends have not occurred. 

C. Effect on Party or Proceeding 

The next element is the effect a recanter's original lie has on the party or
proceeding. Again, with regard to this element, most jurisdictions follow the Model
Penal Code, which drafted its language based on New York's Recantation defense.
The Model Penal Code provides that a recantation defense is viable if, besides
satisfying all other requisite elements, the retraction is made "before the
falsification substantially affected the proceeding." [FN117] 

Most other jurisdictions that subscribe to the recantation rule require this element,
[FN118] but some disregard it completely. [FN119] Again, like the term
"proceeding" discussed earlier, [FN120] it is not entirely clear what "substantially
affected the proceeding" means. Does this mean a burden litigants face from having
to hear a witness testify a second time, this time truthfully? Does this mean
irreversible harm, such as a need for a new trial after a witness died? Or perhaps it

means something simpler like the burden of selecting a new jury? To date, there is
little case law to answer these questions. [FN121] Even the Model Penal Code
leaves no indication of what these words mean in its comments. [FN122] 

This effect-on-party-or-proceeding element, while not being clearly defined, has
been advocated by commentators, [FN123] but it has also had its critics. [FN124] 

V. A Case for the Adoption and the Reformulation of the Recantation Defense 

All but the most tyrannical of people believe a judicial proceeding's chief function
is to bring forth truth. Therefore, it is disturbing that the recantation defense is
unavailable in most jurisdictions throughout the United States. It is also
unfortunate that of those states that have adopted recantation defenses, almost all
of them need reformulation. An overhaul of these defenses would cure one of the
two prevailing problems: first, the defense is inept at accomplishing its function of
encouraging recantations; or secondly, it needs improvement to fully maximize its
truth-enticing potential while discouraging lying. 

A. More Jurisdictions Should Adopt a Recantation Defense 

Although the recantation doctrine has been slowly gaining acceptance, the
majority of states still remain completed-crime jurisdictions. This gives a potentially
repentant witness no way to redeem himself and avoid the peril of a perjury
conviction and, most important, provides no incentive for the witness to speak the
truth after he has lied. 

Completed-crime advocates advance two schools of thought for their rejection of
the recantation doctrine. First, they believe that once the crime is committed, the
"crime is complete" [FN125] - that is, the witness has engaged in culpable behavior
for which he must be punished - and this punishment is deserved from the instant
he utters the lie under oath. This proposition pays homage to retributivism, the
view that society should inflict punishment on a wrongdoer because of his moral
culpability. [FN126] Secondly, completed-crime advocates believe a liar's
punishment serves as both specific and general deterrence to the crime of perjury.
[FN127] Because of this, advocates argue, the completed-crime rule actually
decreases perjury by deterring witnesses from lying when first testifying. 

Arguably, the first of these contentions, retributivism, has some merit in that the
act of lying on the witness stand deserves punishment. It is unnecessary, however,
to engage in the age-old debate on the merits of retributivism to see the
unsoundness of the completed-crime rule. 

If a liar knows the law will punish him for retracting a previously made lie, he will
surely be hesitant to do so. This is especially true in the absence of proof that the
authorities have discovered or will discover his lie. [FN128] Instead, he will most
probably keep his lie a secret. The result of this is terribly ironic: Completed-crime
advocates will fail to accomplish their retributivist goal of punishing the perjurer
because the lie will never likely be discovered in the absence of a recantation
defense. 

The second school of thought is that the completed-crime rule's deterrence value
will decrease the incidence of perjury over the recantation rule. In order to disprove
this theory, it is necessary to discuss some basic criminology. Based on empirical
studies, criminologists universally agree that the two strongest factors in deterring
crime are, first, the severity of the penalty and, secondly, the crime's risk of
apprehension and conviction. [FN129] The latter of the two criteria has proven
most effective for deterring crime, but ironically it is the most difficult to implement.
[FN130] Applying these two factors to the recantation rule, it is apparent that the
recantation defense does not decrease deterrence, as completed-crime advocates
claim. This is because a well-formulated recantation defense [FN131] is available to
liars only when there is little, if any, prospect of discovering the lie and therefore
almost no chance of obtaining a conviction. As for a penalty, jurisdictions vary on
punishment, but most are fairly harsh. [FN132] Because, as most criminologists
believe, the risk of being caught and convicted is the most important criterion to
indicate whether one will commit a crime, the deterrence value of the
completed-crime rule remains doubtful at best. 

While the arguments for the completed-crime rule are precarious, the recantation
rule's sound public policy of bringing forth the truth demands its adoption by all
jurisdictions. Even if, for argument's sake, completed-crime proponents' goals of
retributivism and deterrence are obtainable in a completed-crime jurisdiction, public
policy demands that courts take every measure to bring forth the truth. This is true
even at the expense of letting a liar get away with perjury. Although such a witness
deserves punishment for his initial lie, this punishment should not be at the
expense of litigants whose stake in a trial is often great. [FN133] Bringing out the
truth is even more crucial in a criminal trial where life and liberty are on the line. 

This balancing of public-policy interests is known in philosophy as utilitarianism.
[FN134] Put simply, utilitarianism means that ends must justify the means; or, stated
differently, one may do a "wrong" if its ultimate effect is "good" or best for society.
Applying this to the recantation rule, we except the "wrong" of letting a witness's
lies go unpunished if his later recantation provides the better effect of producing
something "good," which is speaking the truth to the court. 

Today, utilitarianism is manifest in much of our jurisprudence. [FN135] Indeed,
much of the policy and reason behind an array of today's legislation is strictly
utilitarian. Interestingly, some states have employed utilitarianism in such a way
that has resulted in a slightly different recantation defense. 

Some states only allow a recantation defense to a witness in a felony or other
high-level case and not to a low-level trial. [FN136] Perhaps the reasoning for this is
the belief that excusing the "wrong" of perjury can only benefit society if a greater
"good" is accomplished; that "good" being a fair trial of a felony or other high-level
crime based on truth. 

Thus, applied to a basic utilitarian balance, these states presumably reason that it
is more important to punish a liar for perjury than it is to improve the chances of the
truth coming to light in an insignificant low-level trial, such as for a speeding
violation. Because perjury is a greater offense than, for instance, speeding, society
should punish the perjurer without giving him a recantation defense, which would
ultimately aid the defendant or prosecution in a speeding violation hearing. This
also includes other low-level trials where crimes or issues less serious than perjury
are being litigated. Some jurisdictions presumably reason that this gives the
greatest benefit to society by punishing a greater crime. 

Although the intent of making a distinction between high-level trials and low-level
ones is noble, the logic is ultimately flawed for the same reason that the goal of
retributivism is impossible in a completed-crime jurisdiction. [FN137] Witnesses in
these low-level trials will simply not recant their testimony when the lie has not
been or will not be discovered. Hence, the goal of punishing a greater crime at the
expense of a lessor one will not occur, as lying witnesses will remain unrepentant
for fear of prosecution. 

B. The Ideal Recantation Defense 

A well-formulated recantation defense increases the likelihood of veracity and,
contrary to what critics believe, has no risk of encouraging dishonesty. But a
poorly formulated defense, as some courts and commentators rightfully fear, will
indeed encourage lying. [FN138] Likewise, a narrowly applied recantation defense,
while not fostering untruthfulness, will lose the benefits of encouraging truthful
witnesses. 

1. Eliminate Ambiguity in the Language of Most Defenses 

The first step to formulating a model recantation statute is to eliminate ambiguity.
Like New York's retraction statute, [FN139] discussed earlier, [FN140] most
recantation defenses on the books today are adulterated with a lack of specificity in
the language of their terms. As a result, a witness considering retracting a
previously made lie is bedeviled by the question, "Is a defense available to me?"
With the recantation defense, as with all criminal statutes, ambiguity should be
avoided like the plague. Penal consequences are too great to be left to the
capricious nature of a judge's interpretation of legislatures' written memorial - the
statute. Therefore, legislatures must expressly define all terms in their recantation
statutes. In the absence of unequivocal language, a potentially repentant witness
will be hesitant to recant because he lacks knowledge of his fate for doing so.
[FN141] 

Some judicial-activist proponents might argue that these ambiguities can be left to
the courts to decipher. Unfortunately, however, because a witness will hesitate to
admit he lied if he is unclear of the recantation defense's availability, he will likely
elect not to recant. This disincentive to truth telling has resulted and will result in a
barrier to case-law development to correct this ambiguity. Empirical evidence of this
judicial inertia can be seen in New York, where no published case has commented
on the vague terms of New York's retraction statute in the thirty years that the
statute has been in existence. Presumably, potential recanters in New York do not
know whether their recantation was in the same "proceeding" or whether it harmed
the party or proceeding, both of which are necessary to invoke the retraction
defense in New York. [FN142] In summary, case law defining vague recantation
statutes will be extremely slow to develop as potential recanters will be hesitant to
use the defense and take their chances in the appeals process. Therefore, it is
imperative that lawmakers overhaul today's recantation statutes to define all terms
within the language of their respective recantation defenses. [FN143] 

2. Motive or Mens Rea Element is Paramount 

The single most important element for an effective recantation defense is motive.
Disregarding motive, as some recantation defense statutes do, is terribly foolish as
it will likely encourage perjury. Witnesses will lie freely, and later if it becomes
manifest that their lie has or will be discovered, recant their testimony. As a result,
this Article advocates the objective-view-of-motive standard, [FN144] which
affords a defense to a liar only if the authorities have not discovered or will not
discover the lie. 

In applying this motive standard, there should be no leniency. This means any
indication that the lie has been or will be discovered, however slight, should
preclude a defense to the liar. If the motive standard is viewed strictly, as this
Article suggests, there is no chance that the availability of a recantation defense
will encourage perjury, as some courts and commentators fear. However, any lesser
standard of motive very well might encourage perjury. 

One may suggest that a recantation defense should employ the good-faith motive
requirement, which allows a defense to a liar who recants before he, himself,
believes that the lie has been or will be discovered. The good-faith exception is an
unwise choice for two reasons. First, it is a difficult task for any court or jury to
determine one's subjective mind. Secondly, and most important, the lie is culpable
conduct that deserves punishment. The only reason for excusing the lie in the first
place is utilitarianism; [FN145] that is, offering the defense contributes to the
greater public policy of fostering truthfulness in judicial proceedings. In the
absence of any possible benefit for doing so, the lie should be punished. For
instance, if the authorities know of the lie, but the liar himself believes his lie will
remain secret for eternity, the court has discovered the lie and the truth will come to
light, despite any recantation. Because of this, there is no benefit for pardoning the
liar from perjury, which is a culpable act. In the absence of any benefit, the lie must
be punished. As the Supreme Court stated in United States v. Norris, the lie is
culpable behavior from the instant it is uttered. [FN146] Thus, the
objective-view-of-motive standard more appropriately obtains the optimal benefit of
fostering truthfulness while not needlessly excusing perjurers whose recantations
fail to offer the court the greater benefit of veracity; after all, such information has
or will become known without a later-repentant liar's recantation. 

Judge Desmond, in his dissenting opinion in Ezaugi, reasoned that "since the
recantation rule's purpose is not to reward or punish the liar but to get the truth into
the record, the perjurer's motive for recanting has nothing to do with it at all."
[FN147] This argument overlooks the law's obligation to punish perjurers when
there is no benefit for pardoning the crime. Moreover, if the liar is to be found out,
presumably the truth will come to light anyway. Thus, contrary to Judge Desmond's
dissent, the "interest of justice" [FN148] will be served equally if the perjurer is
punished because courts still "get the truth into the record." [FN149] 

Lastly, perhaps in the spirit of utilitarianism, some suggest that the recantation
defense be available to witnesses who recant - even with an impure motive - if once
the lie is discovered, additional corrected testimony comes out that would never
have been discovered and benefits the overall proceeding. [FN150] Logically, this
is a sound and well-reasoned proposition, but in practice it may encourage perjury. 

3. Eliminate All Other Requirements 

As long as the objective-view-of-motive standard requirement is strictly
construed, as suggested above, [FN151] legislatures should eliminate all other
elements. It is irrelevant when a liar ultimately recants his misstatements. Therefore,
to impose an arbitrary locus poenitentiae or time period is without purpose. If
without a recantation the truth will never come to light, then it is unimportant how
long the liar waited until he recanted. This is true even if it is after the conclusion of
the proceeding or trial. [FN152] What is only important is that but for the liar's
recantation the truth will never have come to light. An arbitrary, finite locus
poenitentiae neither discourages perjury nor increases recantations. After the time
period has elapsed, perjurers will keep their lies secret. This fails to further justice.
Therefore, the only sensible locus poenitentiae is one that is contingent upon
motive only, which this Article has called a contingent locus poenitentiae. 

The same is true with the effect-on-party-or-proceeding element. No matter how
irreparably harmed the court or its litigants are, it is senseless to disallow the
recantation defense. If the effect-on-party-or-proceeding element is employed in a
recantation defense, witnesses will elect not to recant once it is apparent that a
party or the proceeding has been harmed. Presumably any recantation of false
testimony offers some benefit, however slight, to the parties and proceeding. After
all, if a party or proceeding has been harmed, a recantation can do nothing but
alleviate some, if not all, of the harm. With the effect-on-party-or-proceeding
element included in defenses, witnesses will refrain from recanting once the lie
harms the party or proceeding, and there will be no chance, however slight, of
reducing that harm caused by speaking the truth. Thus, this element, along with all
others except for the objective-view- of-motive element, should be discarded. 

Keeping the foregoing in mind, this Article advocates the following recantation
statute, which embodies all of the suggestions outlined in this Article: 


S Y Recantation. It shall be a defense to perjury if one who
knowingly lies under oath retracts his falsification before it becomes
manifest that the falsification was or would be exposed to the
authorities.

Disregarding the effect-on-party-or-proceeding and locus poenitentiae elements is
perhaps the most controversial proposition of this Article. The few commentators
who discuss the recantation defense disagree with this Article's argument for their
elimination. Instead, they argue the contrary - that these two elements are indeed
necessary. However, closer reading of their writings reveals their arguments are
conclusory. They fail to explain why these two elements are in fact necessary.
[FN153] 

IV. Conclusion 

In order to maximize the truth-gathering function of judicial proceedings, more
legislatures should adopt properly formulated recantation defenses. However, in
adopting them, legislatures must use unequivocal language and formulate them in
with an objective-view-of-motive, which considers the recanter's motive for
retracting to be paramount in deciding whether to award a recantation defense, as
this Article suggests. While motive is important, all other elements presently
employed by most states' defenses should be discarded. Such a formulation will
yield the following benefits: it will increase recantations of lies; discourage perjury;
and, ultimately, it will best serve public policy by pardoning perjurers only when
the greater good results from doing so. 



FOOTNOTES: 
________________ 

[FNa1]. B.A. Syracuse University 1994; J.D. Candidate Touro College, Jacob D.
Fuchsberg Law Center 1998. I wish to thank Professor Thomas A. Schweitzer of the
Touro Law Center for his encouragement and thoughtful advice. In addition,
special thanks are due to Stephen Kunken, Esq., of Commack, New York, whose
research assignment on New York's retraction-defense statute, N.Y. Penal Law §
210.25 (McKinney 1988), spawned my interest in writing this Article. Finally, I wish
to thank reference librarians Jill B. Selden, Esq., Dr. Gerard E. Giannattasio, Esq., and
the rest of the Touro Law library staff for their invaluable research assistance. 

[FN1]. Although the definition of perjury varies slightly from jurisdiction to
jurisdiction, a general definition can be found in Black's Law Dictionary, which
defines it as follows: 

In criminal law, the willful assertion as to a matter of fact, opinion,
belief, or knowledge, made by a witness in a judicial proceeding as
part of his evidence, either upon oath or in any form allowed by law
to be substituted for an oath, whether such evidence is given in
open court, or in an affidavit, or otherwise, such assertion being
material to the issue or point of inquiry and known to such witness
to be false. A false statement knowingly made in a proceeding in a
court of competent jurisdiction or concerning a matter wherein an
affiant is required by law to be sworn as to some matter material to
the issue or point in question.

Black's Law Dictionary 1139 (6th ed. 1990) (citations omitted). Interestingly, most
jurisdictions emphasize the belief element; that is, a testifying witness must believe
his statement, when made, to be false in order to constitute perjury. Therefore,
some courts will convict a declarant for making a statement that he believes to be
false, even though he may have in fact spoken the truth. See Gordon v. State, 147
N.W. 998 (Wis. 1914); 2 Wharton, Wharton's Criminal Law (11th ed. 1912);
Commonwealth v. Miles, 131 S.W. 385 (Ky. 1910). Thus, in a prosecution for giving
alcohol to a Native American (which, in the past, was a crime in Wisconsin),
prosecution for perjury was appropriate where the accused testified that he had not
given whisky to a Native American, and the recipient of the whisky was not a
Native American, but she believed the recipient to be a Native American. Because
at the time the statement was made the witness believed the whisky recipient was a
Native American, the accused was guilty of perjury. See Gordon, 147 N.W. at 998.
See generally Rollin M. Perkins & Ronald N. Boyce, Criminal Law 518-19 (3d ed.
1982); 2 Joel P. Bishop, Bishop on Criminal Law § 1044 c (John M. Zane & Carl
Zollmann eds., 9th ed. 1923). 

[FN2]. See People v. Simpson, No. BA097211 (Cal. Super. Ct., L.A. County 1995). 

[FN3]. See, e.g., Mark Curriden, The Lies Have It, 81 A.B.A. J. 68 (May 1995)
("Judges, lawyers and experts on the court system worry that perjury is being
committed with greater frequency and impunity than ever before."); Lisa C. Harris,
Note, Perjury Defeats Justice, 42 Wayne L. Rev. 1755, 1777 (1996) (stating that the
offering of false testimony has become commonplace in the courts). 

[FN4]. See, e.g., Anthony Salzman, Recantation of Perjured Testimony, 67 J. Crim. L.
& Criminology 273 (1976) ("Witnesses have violated their judicially administered
oaths to tell the whole truth since the beginning of American jurisprudence . . . .");
Luke Owen Pike, History of the Crime of England 123 (1883) ("(O)ur ancestors
perjured themselves with impunity."). See also Brief for Appellant at 54, People v.
Ezaugi, 141 N.E.2d 580 (N.Y. 1957) ("The tendency to lie even under oath is
substantially the same now as it was three centuries ago."). 

[FN5]. The crime of common law perjury has existed since at least the Seventeenth
Century. See United States v. Norris, 300 U.S. 564, 574 (1937). 

[FN6]. See 2 Frederick Pollock & Frederic William Maitland, The History of the
English Law 242 (2d ed. 1911) ("Very ancient law seems to be not quite certain
whether it ought to punish perjury at all. Will it not be interfering with the business
of the gods?"); 3 James F. Stephen, History of the Criminal Law of England 242
(1883) ("The real singularity is, that for several centuries, no trace is to be found of
the punishment of witnesses for perjury."). 

[FN7]. See Harry Hibschman, "You Do Solemnly Swear!" or That Perjury Problem,
24 J. Am. Inst. Crim. L. & Criminology 901, 903 (1934) (arguing that the value of the
oath in preventing witnesses from lying is negligible). 

[FN8]. Id. at 901. However, in our increasingly secular society, an oath's power of
encouraging truthfulness has diminished. Id. As a result, one author has noted the
importance of another trial device better able to elicit the truth: "Cross-examination,
- the rarest, the most useful . . . has always been deemed the surest test of truth and
a better security than the oath." Francis L. Wellman, The Art of Cross-Examination
vi (4th ed., rev. and enlarged 1936) (quoting Cox). 

[FN9]. But see Harris, supra note 3, at 1777 (arguing that current perjury statutes are
ineffective and need to be made harsher, in addition to adding new laws to facilitate
swifter and certain prosecutions for this crime). 

[FN10]. See Harris, supra note 3, at 1759-62. 

[FN11]. Perhaps the only article dedicated to the recantation doctrine exclusively is
Salzman, supra note 4. 

[FN12]. Admittedly, a witnesses recanting intentionally false testimony is a rarity,
but it does occur on occasion. During such an exceptional occasion the recantation
defense plays a pivotal role in ensuring that a court's justice is based, as much as
possible, on truth rather than lies. 

[FN13]. One may argue - though not necessarily prevail - that due process is
compromised for litigants when testifying witnesses do not have a well- formulated
recantation defense at their disposal if they lie but later wish to recant. This is even
more true in a criminal trial where a defendant's life and liberty are on the line rather
than just money in a civil proceeding. Moreover, the argument goes, due process
demands that the truth come to light at the expense of absolving a liar of his crime
of perjury. 
Keep in mind, though, in no way does the lying witness have a constitutional
right to a retraction defense. As this Article later argues, once a lie is made under
oath, the liar has committed a crime, but public policy requires that the crime be
excused in order to increase the chance that truthful testimony will come to light.
See infra notes 145-46 and accompanying text. Therefore, as United States v.
Denison, 663 F.2d 611 (5th Cir. 1981), explained, no right to a recantation defense
exists to the witness, himself. See also Annotation, Recantation As Bar To Perjury
Prosecution Under 18 U.S.C.S. § 1623(d), 65 A.L.R. Fed. 177, 184-86 (1983). But, as
explained above, an argument can be made that not availing a witness the
recantation defense diminishes due process rights of litigants in both civil and,
especially, criminal proceedings. 

[FN14]. See United States v. Norris, 300 U.S. 564, 574 (1937). 

[FN15]. The following are recantation defense statutes that exist in a minority of
jurisdictions: Ala. Code § 13A-10-107 (1995); Alaska Stat. § 11.56.235 (1996); Ark.
Code Ann. § 5-53-104 (Michie 1997); Colo. Rev. Stat. Ann. § 18-8-508 (West 1986);
Del. Code Ann. tit. 11, § 1231 (1995); Fla. Stat. Ann. § 837.07 (West 1994); Haw.
Rev. Stat. § 710-1064 (1993); Ill. Comp. Stat. Ann. ch. 720, para. 5/32-2 (c) (West
1993); Iowa Code Ann. § 720.2 (West 1993); Ky. Rev. Stat. Ann. § 523.090
(Michie/Bobbs- Merrill 1990); Me. Rev. Stat. Ann. tit. 17-A, § 451(3) (West 1983);
Mont. Code Ann. § 45-7-201(5) (1997); N.J. Stat. Ann. § 2C:28-1(d) (West 1995);
N.Y. Penal Law § 210.25 (McKinney 1988); N.D. Cent. Code § 12.1-11- 04(3) (1997);
18 Pa. Cons. Stat. Ann. § 4902(d) (1983); R.I. Gen. Laws § 11-33-1(d) (1994); Tex.
Penal Code Ann. § 37.05 (West 1994); Wash. Rev. Code Ann. § 9A.72.060 (West
1988). 
The federal government adopted the recantation rule, which it codified in 18
U.S.C. § 1623(d) (1994). Section 1623(d) affords a recantation defense to statements
made under oath only before a grand jury or court. On the other hand, 18 U.S.C. §
1621 (1994), which is applicable more generally to any statement given under oath,
disallows the retraction defense. The disparity between these two statutes makes it,
at times, unclear if a liar may invoke a recantation defense. As such, the federal
retraction defense has drawn criticism from many commentators. See infra note 141. 

[FN16]. See 2 Sara S. Beale & William C. Bryson, Grand Jury Law and Practice §
11:10 (1986) (noting that "making the recantation defense too broad can have the
opposite effect of encouraging perjury"); Salzman, supra note 4, at 279; Recent
Case, Criminal Law - Perjury - Correction of False Testimony, 76 U. Pa. L. Rev. 751,
752 (1927). See, e.g., United States v. Norris, 300 U.S. 564, 574 (1937). 

[FN17]. N.Y. Penal Law § 210.25 (McKinney 1988) (effective Sept. 1, 1967, codified
1965). 

[FN18]. Cf. King v. Jones, 1 Peake 51, 53 (1791) (citing King v. Carr, 82 Eng. Rep.
1191 (1669)). But see United States v. Norris, 300 U.S. 564 (1937) (citing Edwards v.
M'Leay, 35 Eng. Rep. 316 (1813); Reg. v. Holl, 45 L.T.R. 69 (Q.B.D. 1881)) (arguing
that there is doubt that Carr held or intended to hold that a witness's retraction of
his false testimony absolves him of perjury in light of later English case law). 

[FN19]. 111 N.Y.S. 133 (N.Y. App. Div. 1908). 

[FN20]. See id. at 134. Mr. Gillette was subpoenaed to testify before a grand jury in
a proceeding entitled "The People of the State of New York v. John Doe et al." Id. 

[FN21]. Id. 

[FN22]. Id. at 138. 

[FN23]. Id. at 139. 

[FN24]. For the general definition of perjury see supra note 1. 

[FN25]. King v. Jones, 1 Peake's Reports 51 (1791) (citing King v. Carr, 82 Eng. Rep.
1191 (1669)). The Carr case, which Jones cited, was written in law French as was the
practice in England at that time period. 

[FN26]. Id. at 53. 

[FN27]. See, e.g., Reg. v. Holl, 45 L.T.R. 69, 70 (Q.B.D. 1881) ("(a)n indictment for
perjury could not be sustained on an answer afterwards corrected or explained."). 

[FN28]. See, e.g., Henry v. Hamilton, 7 Blackf. 506, 507 (Ind. 1845) (approving a trial
court's instruction that a witness' corrected statement may be considered to negate
the willfulness element necessary for a perjury conviction). 

[FN29]. See Model Penal Code § 241.1cmt. 7, 130-31 (1980) ("Under prevailing law . .
. a prompt retraction . . . (can be used) to bolster the assertion that the original
misstatement was inadvertent or due to a misunderstanding."); Salzman, supra note
4, at 275 ("(C)ourts generally agree that an offer of testimonial correction is relevant
to show that the inaccurate testimony was not deliberately false and that no perjury
was therefore ever committed."). 

[FN30]. See People v. Gillette, 111 N.Y.S. 133, 139 (N.Y. App. Div. 1908). 

[FN31]. Id. 

[FN32]. See id. 

[FN33]. Id. 

[FN34]. See id. 

[FN35]. 141 N.E.2d 580 (N.Y. 1957). 

[FN36]. See id. at 582. 

[FN37]. Id. 

[FN38]. Id. 

[FN39]. Id. at 582 & n.1. 

[FN40]. Id. at 582. 

[FN41]. Id. 

[FN42]. Id. 

[FN43]. Id. 

[FN44]. Id. 

[FN45]. 300 U.S. 564 (1937). 

[FN46]. Although a federal decision on perjury recantation is not binding on state
courts, Detective Ezaugi's attorneys feared the Ezaugi court would find Norris
persuasive and, thus, argued vehemently against New York adopting it: 

[E]Ven if the Federal Courts limit the dectrine [sic] of recantation as
last enunciated in People v. Gillette [ ] there is no reason for this
Court to renounce the Gillette case. . . . Even if the Norris rule
completely and without exception discredited the Gillette rule, which
it does not, it would not be the first time that the United States rule
of policy was different from the state rule in a particular instance. A
most notable illustration is that in the United States Courts, a
constitutional prohibition against unlawful search and seizure is
rigidly observed, whereas in our Courts we do not have the
enforcement of such a prohibition despite a similar state
constitutional provision.

Brief for Appellant at 53-54, People v. Ezaugi, 141 N.E.2d 580 (N.Y. 1957) (citing
People v. Defore, 150 N.E. 585 (N.Y. 1926), cert. denied, 270 U.S. 657 (1926)). 

[FN47]. See infra note 65 and accompanying text. 

[FN48]. Norris, 300 U.S. at 576 (emphasis added). 

[FN49]. See id. at 574. 

[FN50]. Ezaugi, 141 N.E.2d at 583. 

[FN51]. Id. 

[FN52]. See id. 

[FN53]. Id. at 583 (Desmond, J., dissenting) (stating that the new elements, "while
appearing to reaffirm the ancient and sound recantation rule, (citation omitted)
actually so limits and hedges that rule as to leave it without any utility"). 

[FN54]. See New York State Commission on Revision of the Penal Law and Criminal
Code, Proposed New York Penal Law, Commission Staff Notes 135 (1964). 

[FN55]. Model Penal Code § 241.1(4) (1967). See infra note 116 and accompanying
text for the full text of this statute. 

[FN56]. See Ezaugi, 141 N.E.2d at 583. 

[FN57]. See N.Y. Penal Law § 210.25 (McKinney 1988). 

[FN58]. The retraction statute is equally applicable to civil trials, but it may have
stronger due process implications for criminal trials. See supra note 13. 

[FN59]. For a more detailed discussion of locus poenitentiae, see infra notes 96-116
and accompanying text. 

[FN60]. People v. Brill, 165 N.Y.S. 65, 71 (N.Y. Ct. of Gen. Sessions, N.Y. County
1917) (citing People v. Gillette, 111 N.Y.S. 133 (N.Y. App. Div. 1908)). 

[FN61]. See People v. Markan, 206 N.Y.S. 197, 199 (N.Y. Ct. of Gen. Sessions, N.Y.
County 1924). 

[FN62]. See People v. Ashby, 195 N.Y.S.2d 301, 304 (N.Y. App. Div. 1959), rev'd on
other grounds, 168 N.E.2d 672 (N.Y. 1960). 

[FN63]. N.Y. Penal Law § 210.25, Arnold D. Hechtman, Practice Commentaries, 488,
489-90 (McKinney 1975); N.Y. Penal Law § 210.25, Richard G. Denzer & Peter
McQuillan, Practice Commentary, 710, 712 (McKinney 1967). The preceding
commentaries, by different authors writing on the same statute, are identical.
Interestingly, the subsequent McKinney commentary included in N.Y. Penal Law §
210.25, William C. Donnino, Practice Commentaries, 515, 521 (McKinney 1988),
makes no mention of this distinction. Perhaps the last author, Mr. Donnino,
because of a lack of controlling authority, disbelieves the distinction asserted by
his predecessors to the McKinney commentaries. 

[FN64]. See infra notes 104-07 and accompanying text. 

[FN65]. 300 U.S. 564, 574 (1937). 

[FN66]. See id. Keep in mind, the key elements for perjury culpability are that the
false statement was known to be false and made willfully. Without these elements,
perjury cannot be predicated. But, also remember, a witness threatened with perjury
can argue that a subsequent statement to correct or clarify previous testimony is
indicative that one did not willfully and knowingly lie from the start. In such an
instance, a crime has not been committed. See supra notes 1, 22-29 and
accompanying text. 

[FN67]. Norris, 300 U.S. at 574. 

[FN68]. See id. at 574. See also Loubriel v. United States, 9 F.2d 807 (2d Cir. 1926);
Martin v. Miller, 4 Mo. 39 (1835). 

[FN69]. See Norris, 300 U.S. at 574. 

[FN70]. See W.M. Moldoff, Annotation, Recantation as Defense in Perjury
Prosecution, 64 A.L.R.2d 276, at 278 (1959). 

[FN71]. See 18 U.S.C. § 1623 (1994). 

[FN72]. See Wis. Stat. § 946.31 (1996). 

[FN73]. Today, almost all recantation defenses are statutory. 

[FN74]. The American Law Institute codified the Ezaugi decision in Model Penal
Code § 241.1(4) (1967), changing the language slightly, which, in turn, New York's
legislature adopted in its codification of the recantation defense in N.Y. Penal Law §
210.25 (McKinney 1988). Since then, most states with recantation defenses have
adopted the Ezaugi decision, as enunciated in the Model Penal Code's language. 

[FN75]. See Norris v. United States, 86 F.2d 379 (8th Cir. 1936), rev'd, 300 U.S. 564
(1937); Salzman, supra note 4, at 280; W. M. Moldoff, Annotation, Recantation as
Defense in Perjury Prosecution, 64 A.L.R.2d 276, at 278 (1959). 

[FN76]. Black's Law Dictionary 1014 (6th ed. 1990). 

[FN77]. 111 N.Y.S. 133 (N.Y. App. Div. 1908). 

[FN78]. See United States v. Norris, 300 U.S. 564, 575 (1937) (criticizing Gillette's
precedential value because it was not rendered by the court of appeals, New York's
highest court, and because a subsequent case, People v. Markan, 206 N.Y.S. 197
(N.Y. Ct. of Gen. Sessions, N.Y. County 1924), refused to follow Gillette where a
contradictory statement was not part of the same examination at which the first
statement was uttered); People v. Ezaugi, 141 N.E.2d 580, 582-83 (N.Y. 1957)
(implying that the recantation defense should not be universally applied in
situations where the liar has a tainted motive for recanting). 

[FN79]. See, e.g., Colo. Rev. Stat. Ann. § 18-8-508 (West 1986); N.J. Stat. Ann. §
2C:28-1 (d) (West 1995). 

[FN80]. Model Penal Code § 241.1 (4) (1997). See also the following state
recantation statutes, which use the same or similar language: Ala. Code §
12A-10-107 (1995); Del. Code Ann. tit 11, § 1231 (1996); Fla. Stat. Ann. § 837.07
(West 1994); Haw. Rev. Stat. § 710-1064 (1993); Ill. Comp. Stat. Ann. ch. 720, para.
5/32-2 (c) (West 1993); Ky. Rev. Stat. Ann § 523.090 (Michie/Bobbs-Merrill 1990);
Me. Rev. Stat. Ann. tit. 17-A, § 451 (3) (West 1983); Mont. Code Ann. § 45-7-201(5)
(1997); N.D. Cent. Code § 12.1-11-04 (3) (1997); N.J. Stat. Ann. § 2C:28-1 (d) (West
1995); N.Y. Penal Law § 210.25 (McKinney 1988); Or. Rev. Stat. § 162.105 (1995); 18
Pa. Cons. Stat. Ann. § 4902 (d) (West 1983); R.I. Gen. Laws § 11-33-1 (d) (1994);
Wash. Rev. Code Ann. § 9A.72.060 (West 1988). 

[FN81]. See United States v. Clavey, 578 F.2d 1219, 1222 n.5 (7th Cir. 1978); State v.
Hanson, 302 N.W.2d 399 (N.D. 1981). Both the Clavey and Hanson cases hold that
their recantation statutes are based on N.Y. Penal Law § 210.25 (McKinney 1988),
which is based on the rule enunciated in People v. Ezaugi, 141 N.E.2d 580 (N.Y.
1957). One part of the original Ezaugi rule, the Hanson court pointed out, is that a
successful retraction defense is viable "when no reasonable likelihood exists that
the witness has learned that his perjury is known or may become known to the
authorities." Hanson, 302 N.W.2d at 403 (quoting Ezaugi, 141 N.E.2d at 583)
(emphasis added). This, essentially, is the good-faith standard, which is discussed
later. See infra 87-88 and accompanying text. Thus, the Hanson court believes 
that the exposure of the perjury becomes "manifest" when the defendant knows or
has reason to know that the authorities are or will be aware of the falsification . . . . it
may be important to know whether or not the authorities have already discovered,
or are certain to discover, the falsification when we are assessing the defendant's
state of mind, but that alone does not determine the validity of the retraction
defense. It is not the state of mind of the authorities that controls. 
Hanson, 302 N.W.2d at 403. 
The federal judiciary shares similar sentiments and interprets this language as
applying to the witness, himself. See, e.g., United States v. Denison, 663 F.2d 611
(5th Cir 1981); United States v. Serimgeour, 636 F.2d 1019 (5th Cir. 1981); United
States v. Tucker, 495 F. Supp. 607 (E.D.N.Y. 1980); United States v. Swainson, 548
F.2d 657 (6th Cir. 1977); United States v. Mazzei, 400 F. Supp. 17 (W.D. Pa. 1975).
See generally Annotation, 65 A.L.R. Fed. 177, 191-95 (1983). 

[FN82]. See infra notes 87-88 and accompanying text where this Article discusses
the good-faith motive. 

[FN83]. This Article later advocates this type of motive standard for an ideal
recantation defense. See infra notes 138-53 and accompanying text. 

[FN84]. The element would also be fulfilled if at the time of the trial for perjury it
comes to light that the lying subsequently came to the attention of the authorities. 

[FN85]. Again, we presume that this language applies to the liar's subjective belief
that the authorities have or will discover his lies. There is little, if any, case law to
contradict this hypothesis. Assuming, however, this language is inapplicable to the
liar's mens rea, there would be a different outcome under the following hypothetical:
A witness, who happens to be nervous and perhaps a bit paranoid, convincingly
lies on the witness stand. The witness subjectively believes the authorities have or
will discover his falsehoods, but in reality there is no prospect of the disclosure.
Because of this fear, he says to the court, "I know the prosecution learned my
testimony was intentionally false; therefore, I would like to retract it now and
replace it with the truth," which he does. If we interpret the language as not
applying to the liar's beliefs, the witness will have successfully fulfilled the
objective- view-of-motive element under this interpretation of the language, and he
would be protected from perjury prosecution. (Only if the authorities have or will
possibly catch the lie can the witness be prosecuted.) It would be irrelevant that his
mens rea is guilty, in that he believes he will soon be caught. 

[FN86]. See supra note 85. 

[FN87]. The language quoted above is fictitious; no statute uses precisely the same
language, but the following statutes have similar language: Alaska Stat. §
11.56.235(b)(1) (1996); Haw. Rev. Stat. § 710-1064(1)(a) (1993). In addition, Or. Rev.
Stat. § 162.05(a) (1997) has similar motive language, which provides a retraction to
be made "in a manner showing complete and voluntary retraction of the prior false
statement." Id. 

[FN88]. See supra notes 80-85 and accompanying text. 

[FN89]. Colo. Rev. Stat. Ann. § 18-8-508 (West 1986); Ill. Comp. Stat. Ann. ch. 720,
para. 5/32-2 (c) (West 1993); Iowa Code Ann. § 720.2 (West 1993). See also People
v. Ezaugi, 141 N.E.2d 580, 584 (N.Y. 1957) (Desmond, J., dissenting) (arguing that
motive element should not be considered for deciding the availability of the
recantation defense); Commonwealth v. Irvine, 14 Pa. D. & C. 275 (1930). Today, the
Irvine court's adoption of the no-motive standard is not followed because
Pennsylvania's retraction statute, 18 Pa. Cons. Stat. Ann. § 4902(d) (1983),
specifically provides for a motive requirement. 

[FN90]. See Ezaugi, 141 N.E.2d at 582-83. In addition, this Article is critical of the
no-motive standard. 

[FN91]. See id. at 583; Salzman, supra note 4, at 280. 

[FN92]. United States v. Norris, 300 U.S. 564, 574 (1937); Recent Case, Criminal Law
- Perjury - Retraction of False Testimony Held No Bar to Prosecution, 51 Harv. L.
Rev. 165 (1937) ("Since, however, a perjurer will not usually retract unless his
falsehood has been demonstrated, retractions thus induced will be of little value in
furthering the administration of justice."). 

[FN93]. Ezaugi, 141 N.E.2d at 584 (Desmond, J., dissenting). 

[FN94]. See Model Penal Code § 241 (4) (1962). This code section provides that a
retraction must be made "before it became manifest that the falsification was or
would be exposed." Id. 

[FN95]. Model Penal Code § 241 cmt. 7 (1980). 

[FN96]. Justice Roberts in United States v. Norris, 300 U.S. 564, 572 (1937), uses
this term to describe the time period in which one has to recant his false statement. 

[FN97]. Ezaugi, 141 N.E.2d at 580. 

[FN98]. Model Penal Code § 240.1 (4) (1962). See infra note 116 and accompanying
text for the full text. 

[FN99]. N.Y. Penal Law § 210.25 (McKinney 1988). 

[FN100]. Ezaugi, 141 N.E.2d at 583. 

[FN101]. Ill. Comp. Stat. Ann. ch. 720, para. 5/32-2 (c) (West 1993). See also Fla. Stat.
Ann. § 837.07 (West 1996) (using the language "in the same continuous proceeding
or matter"). 

[FN102]. Tex. Penal Code Ann. § 37.05 (West 1995). 

[FN103]. Or. Rev. Stat. § 162.105(c) (1997). 

[FN104]. See supra notes 59-64 and accompanying text. 

[FN105]. State in the Interest of J.S., 642 A.2d 430 (N.J. Super. Ct. Ch. Div. 1994). 

[FN106]. See People v. Valdez, 568 P.2d 71 (Colo. 1977). Although Colorado's
statute codifying the recantation defense already had defined the word
"proceeding" by statute, the Valdez court further defined it as including various
stages of a trial, but not a mistrial. Id. 

[FN107]. Ark. Code Ann. § 5-53-104 (Michie 1995) provides: "Statements made in
separate hearings at separate stages of any official proceeding shall be deemed to
have been made in the course of the same proceeding." But this is limited by Brown
v. State, 707 S.W.2d 313 (1986), which held in the context of Arkansas' recantation
statute, that a hearing plea withdrawal and the hearing on the accepted guilty pleas
were not part of the same "proceeding" when the previous phase ended. Colo. Rev.
Stat. Ann. § 18-8-508 (West 1986) in part, provides: "Statements made in separate
hearings at separate stages of the same trial or administrative proceeding shall be
deemed to have been made in the course of the same proceeding." Id. 

[FN108]. Although there may be language defining the period of time, it is often not
clear what that language means, unless there is adequate case law to explain it
further. 

[FN109]. Colo. Rev. Stat. Ann. § 18-8-508 (West 1990). 

[FN110]. Id. Although Colorado defines "procedure" within the context of its
retraction defense, many other states have not developed a specific definition. 

[FN111]. Although most the locus poenitentiae for most defenses have an event
upon which it is contingent, the vast majority are hybrid locus poenitentiae. In
other words, they require that one recant before a contingent event (e.g., before the
parties and procedure are prejudiced), in addition to having to follow a finite period
of time (e.g., before the conclusion of the proceeding). 

[FN112]. The following are some defenses that require one to recant before one or
both of these contingencies: Ala. Code § 12A-10-107 (1995); Del. Code Ann. tit 11,
§ 1231 (1995); Fla. Stat. Ann. § 837.07 (West 1994); Haw. Rev. Stat. § 710-1064
(1993); Ill. Comp. Stat. Ann. ch. 720, para. 5/32-2 (c) (West 1993); Ky. Rev. Stat. Ann
§ 523.090 (Michie 1990); Me. Rev. Stat. Ann. tit. 17-A, § 451 (3) (West 1983); Mont.
Code Ann. § 45-7- 201(5) (1997); N.D. Cent. Code § 12.1-11-04(3) (1997); N.J. Stat.
Ann. § 2C:28-1(d) (West 1995); N.Y. Penal Law § 210.25 (McKinney 1988); Or. Rev.
Stat. § 162.105 (1997); 18 Pa. Cons. Stat. Ann. § 4902(d) (1983); R.I. Gen. Laws §
11-33-1 (d)(1994); Wash. Rev. Code Ann. § 9A.72.060 (West 1988). 

[FN113]. Section X is a fictitious statute. At present, no jurisdictions employ a
solely contingent locus poenitentiae. (Most employ one that is both contingent
and dependent on a fixed time period, which this Article calls a hybrid locus
poenitentiae.) This Article later advocates an ideal reformulation of recantation
statutes and employs the contingent locus poenitentiae type of locus potentate.
See infra notes 151-53 and accompanying text. 

[FN114]. Salzman, supra note 4, at 279-280. 

[FN115]. See People v. Ezaugi, 141 N.E.2d 580, 583 (N.Y. 1957). 

[FN116]. Model Penal Code § 240.1 (4) (1962). 

[FN117]. Id. 

[FN118]. See Ala. Code § 13A-10-107 (1995); Del. Code Ann. tit. 11, § 1231 (1995);
Fla. Stat. Ann. § 837.07 (West 1994); Ky. Rev. Stat. Ann. § 523.090 (Michies 1990);
Mont. Code Ann. § 45-7-201(5) (1997); N.Y. Penal Law § 210.25 (McKinney 1988);
N.D. Cent. Code § 12.1-11-04(3) (1997); 18 Pa. Cons. Stat. Ann. § 4902(d) (West
1983); Wash. Rev. Code Ann. § 9A.72.060 (West 1988). 

[FN119]. See, e.g., Colo. Rev. Stat. Ann. § 18-8-508 (West 1986); Ill. Comp. Stat.
Ann. ch. 720, para. 5/32-2 (c) (West 1993); Me. Rev. Stat. Ann. tit. 17-A, § 451 (3)
(West 1983); Tex. Penal Code Ann. § 37.05 (West 1994). See also N.J. Stat. Ann. §
2C:28-1 (d) (West 1995), which provides similar language: "(W)ithout having
caused irreparable harm to any party." Arkansas provides "any person who in
making a retraction causes termination of any official proceeding by reason of
prejudice to a legal right of party to the proceeding shall be guilty of a Class A
misdemeanor." Ark. Code Ann. § 5-53- 104 (Michie 1997). 

[FN120]. See supra notes 17-64 and accompanying text. 

[FN121]. But see, e.g., United States v. Anfield, 539 F.2d 674 (9th Cir. 1976); United
States v. Slawick, 408 F. Supp. 190 (D.C. Del 1975); United States v. Crandall, 363 F.
Supp. 648 (W.D. Pa. 1973); United States v. Krogh, 366 F. Supp. 1255 (D.C. Dist.
1973); Annotation, 65 A.L.R. Fed. 177, 189-91 (1983). 

[FN122]. See Model Penal Code § 241.1 cmt. 7 (1980). 

[FN123]. Harris, supra note 3, at 1792. 

[FN124]. Ala. Code § 12A-10-107 commentary (1996) (arguing that the vagueness
on what "substantially affects a proceeding is undesirable"). 

[FN125]. United States v. Norris, 300 U.S. 564, 574 (1937). 

[FN126]. See generally Joshua Dressler, Understanding Criminal Law S2.03 (c)
(1987); Wayne R. LaFave & Austin W. Scott, Jr., Criminal Law § 1.5 (a) (6) (2d ed.
1986); Immanuel Kant, The Metaphysical Elements of Justice 99- 107 (J. Ladd trans.,
1965). 

[FN127]. For information on specific and general deterrence see generally Sanford
H. Kadish & Stephen J. Schulhofer, Criminal Law and Its Processes: Cases and
Materials 115 (6th ed. 1995). 

[FN128]. See infra notes 138-50 and accompanying text where this Article discusses
this factor in greater detail, in addition to advocating that motive for recanting be
the most important element of a recantation defense. 

[FN129]. See Franklin Zimring & Gordon Hawkins, Deterrence: The Legal Threat in
Crime Control 158-72 (1973). 

[FN130]. See generally Kaddish & Schulhofer, supra note 127, at 101-31. 

[FN131]. See supra notes 128-53 and accompanying text. 

[FN132]. See, e.g., Mich. Comp. Laws Ann. § 750.423 (West 1996) (providing a
felony punishment not more than fifteen years in state prison). 

[FN133]. Cf. Bussey v. State, 64 S.W. 268, 269 (Ark. 1901). 

[FN134]. Classical utilitarianism was formulated over two centuries ago by Jeremy
Bentham. See Jeremy Bentham, An Introduction to the Principles of Morals and
Legislation (1789). See generally Dressler, supra note 126, § 2.03. 

[FN135]. See, e.g., H.L. Pohlman, Justice Oliver Wendell Holmes & Utilitarian
Jurisprudence (1984). 

[FN136]. 8 See, e.g., Colo. Rev. Stat. Ann. § 18-8-508 (West 1996) (retraction
defense available only against charges of perjury in the first degree); Del. Code
Ann. tit 11, § 1231 (1995) (retraction defense available only for perjury, not for
misdemeanor of making a false written statement); Tex. Penal Code Ann. § 37.05
(West 1994) (retraction defense only available for felony of aggravated perjury, not
for misdemeanor of simple perjury). 

[FN137]. See supra note 126 and accompanying text. 

[FN138]. See United States v. Denison, 663 F.2d 611 (5th Cir. 1981). 

[FN139]. N.Y. Penal Law § 210.25 (McKinney 1988). 

[FN140]. See supra notes 54-57 and accompanying text. 

[FN141]. The retraction provision of the federal perjury statute, 18 U.S.C. § 1623(d)
(1994), has received criticism because of the doubt over whether it or 18 U.S.C. §
1621 (1994), another perjury statute that disallows the retraction defense, is
applicable. Although both are perjury statutes, only the former contains a
recantation defense in sub-part (d). Thus, if unsure which statute is applicable to
them, witnesses will likely choose not to recant. For a detailed discussion of this
see George W. Aycock, III, Note, Nothing But the Truth: A Solution to the Current
Inadequacies of the Federal Perjury Statutes, 28 Val. U. L. Rev. 247 (1993). See
generally Harris, supra note 3, at 1792; 65 Am. Jur. 2d Perjury § 107 (1988); Susan
W. Brenner & Gregory G. Lockhart, Federal Grand Jury Practice S13.17 (1993); The
Grand Jury Project Inc. of the National Lawyers Guild, Representation of Witnesses
Before Federal Grand Juries § 15.3 (d) (3d ed. 1993); Salzman, supra note 4, at
280-86. 

[FN142]. The ambiguity of New York's recantation defense was discussed in detail
earlier above. See supra notes 58-64 and accompanying text. 

[FN143]. As mentioned earlier, some state statues have defined important terms
within their statutes, and this is indeed wise. See supra note 107. More jurisdictions
should do so as well. 

[FN144]. See supra notes 80-86, where the objective-view-on-motive element is
discussed in detail. 

[FN145]. See supra notes 134-35 and accompanying text. 

[FN146]. 300 U.S. 564, 574 (1937). 

[FN147]. People v. Ezaugi, 141 N.E.2d 580, 584 (N.Y. 1957) (Desmond, J., dissenting).

[FN148]. Id. 

[FN149]. Id. 

[FN150]. United States v. Del Toro, 513 F.2d 656 (2d Cir. 1975); Salzman, supra note
4, at 280. 

[FN151]. See supra notes 144-50 and accompanying text. 

[FN152]. Recantation of perjured testimony produces a complex set of issues in
both criminal and civil trials involving how or whether to go about a retrial. In the
interest of brevity, this Article does not touch upon them. For more information, see
generally Janice J. Repka, Comment, Rethinking the Standard for New Trial Motions
Based upon Recantations of Newly Discovered Evidence, 134 U. Pa. L. Rev. 1433
(1986); Sharon Cobb, Comment, Gary Dotson as Victim: The Legal Response to
Recanting Testimony, 35 Emory L.J. 969 (1986). 

[FN153]. Harris, supra note 3, at 1792. See also Salzman, supra note 4, at 280. Mr.
Salzman, contrary to the previously mentioned commentator, Ms. Harris, offers at
least some support for the effect-on-party-or-proceeding and locus poenitentiae
elements as articulated in People v. Ezaugi, 141 N.E.2d 580 (N.Y. 1957).
Nevertheless, as this Article has illustrated, his arguments for
effect-on-party-or-proceeding and locus poenitentiae elements lack merit. Mr.
Salzman writes: 
While some state and federal courts have frustrated the development of a
workable recantation rule through misplaced emphasis on chronological timeliness
(which this Article calls a fixed locus Poenitentiae), the "New York rule" (see supra
note 75 and accompanying text) advanced in Ezaugi recognized the proper
interweaving of the factors of timeliness, motive, and prejudice as a better solution
to the perjury problem. The Ezaugi test requires that the presiding judge inquire
initially into the "timeliness" of a correction; however, the question of timeliness
turns on whether the testimony sought to be corrected has already prejudiced the
administration of justice and on whether the witness believed he was providing
information with his correction which the authorities did not already have. By
varying the intensity of this scrutiny, a deciding court can encourage corrections
whenever it deems correction helpful without making the privilege available to
every potential perjurer. 
Salzman, supra note 4, at 280 (citation omitted). Additionally, Mr. Salzman
advocates - in a more conclusory manner - that recantation defenses should
incorporate the effect-on-party-or-proceeding element: "Extreme cases, such as a
correction offered subsequent to the completion of a trial, surely cannot be
permitted. Thus, even under the most liberal view of 'immediately,' there must be
some point after which a correction will always be too late." Id. at 279 n.54. Notice
that Mr. Salzman fails to give any reason for such a time limitation.