OPINION
KEASLER, J.,
delivered the opinion of the Court,
in which McCORMICK, P. J., and MEYERS, MANSFIELD, PRICE, HOLLAND, and JOHNSON, J.J., joined.
This case presents a complicated time-credit issue, specifically, how to compute time on a stacked sentence.
I. Facts
Greg Kuester was initially sentenced to a ten-year term in the Texas Department of Criminal Justice, Institutional Division (TDCJ-ID), for burglary of a building committed in 1988. On September 1,1989, while in the custody of TDCJ-ID, Kuester struck a correctional officer. Kuester was convicted of aggravated assault on July 25, 1991. The trial court sentenced Kuester to a four-year term in prison and ordered that the sentence would not begin to run until Kuester’s ten-year sentence “ceased to operate.”
Upon receipt of the new conviction information, TDCJ-ID recalculated Kuester’s records, adding the new four-year sentence to increase his prison term from ten to fourteen years. Kuester remained in prison, serving his combined sentences, until September 8, 1992, when he was released to parole. He was at liberty until he was arrested pursuant to a parole violator warrant on August 29, 1993. The Board of Pardons and Paroles (the Board) voted to revoke his parole, and he was returned to the custody of TDCJ-ID to continue serving his combined sentences.
In 1997, TDCJ and the Board determined that time calculations for consecutively-sentenced inmates required re-evaluation for offenses committed after 1987.1 In particular, the practice of treating cumulative sentences as a single combined sentence was discontinued in lieu of performing time calculations on each consecutive case singularly and sequentially.2 Accordingly, TDCJ-ID no longer permitted inmates to accrue time credit for a consecutive sentence until the prior sentence was determined to have ceased to operate, either by day-for-day discharge or upon a parole approval vote by the Board.
In accord with revised policies, Kues-ter’s records were reprocessed to reflect confinement under only the initial ten-year sentence and that the four-year sentence had not yet begun. On May 2, 1999, TDCJ-ID and the Board determined that Kuester’s first sentence ceased to operate. At that time, Kuester began serving his aggravated assault case, designated to discharge on April 27, 2003. This date does not appear to give Kuester credit for pretrial jail time, time he served in prison before his initial release to parole, time he [266]*266spent while released to parole, or much of the time he has served since his parole revocation.
Kuester contends that his release to parole in 1992 was erroneous because TDCJ-ID failed to treat his two sentences separately. He seeks additional credit to his aggravated assault sentence for the period of his erroneous release. We begin by analyzing the applicable statute.
II. Art. 42.08(b) — Completion of the sentence
Article 42.083 sets out the manner in which consecutive sentences are to be carried out. Subsection (a) addresses cases in which the trial court, in its discretion, stacks sentences. Subsection (b) addresses cases in which the trial court is required to stack sentences, specifically, cases in which an offense is committed while the defendant is already an inmate in TDCJ-ID. The two subsections use different phrases to describe the point at which the cumulated sentence begins. Subsection (a) states that the second sentence should begin when the first sentence “has ceased to operate.” Subsection (b), on the other hand, states that the second sentence should begin “on completion of the [first] sentence.” We must determine what the phrase “completion of the sentence” means and, in particular, whether it means something different than “cease to operate.”
A. Language of Stacking Order
Initially, we note that the trial court ordered that Kuester’s aggravated assault sentence begin to run after his ten-year sentence “ceased to operate.” This language appears in Subsection (a). But
Kuester’s aggravated assault conviction was committed while Kuester was an inmate of TDCJ-ID, so the trial judge was required to cumulate the sentence pursuant to Subsection (b) rather than Subsection (a). Nevertheless, regardless of the language in the judgment, the statute dictates the proper application of the order.4 Therefore, we must determine when Kues-ter’s initial sentence was “completed” so that his subsequent sentence could begin.
B. Plain Language
The starting point for statutory analysis is the text of the relevant provisions.5 We first look to the plain meaning of the words to determine the purpose and effect of a statute. We may only look beyond the text if the language is ambiguous or would lead to an absurd result that the Legislature could not have possibly intended.6
The text of Subsection (b) has undergone only minor revisions since the date of Kuester’s offense. It currently provides as follows:
If a defendant is sentenced for an offense committed while the defendant was an inmate in the institutional division of the Texas Department of Criminal Justice and the defendant has not completed the sentence he was serving at the time of the offense, the judge shall order the sentence for the subsequent offense to commence immediately on completion of the sentence for the original offense.7
We construe words and phrases according to the rules of grammar and common usage unless they have acquired, a technical or particular meaning.8 Webster [267]*267defines “complete” as entire, total, undivided, and implies that a certain unit has all of its parts fully developed, the process being carried to its fulfillment.9 So common understanding of “complete” in this context means that the sentence must be wholly discharged; release to parole would not constitute “completion” of the sentence.
Applying “discharge-in-full” as the plain meaning of “completion” results in two potential consequences. One possibility is that an inmate makes parole on his first sentence. He has not discharged his first sentence, so his second sentence cannot yet begin. His second sentence will not begin until his first sentence has discharged, that is, until his parole is complete. The inmate may spend twenty or thirty years on parole, living and working in the community, a rehabilitated and contributing member of society. When at last his parole is complete and his sentence discharges, he must then be returned to prison to begin serving his second sentence. Or, in the case of a particularly long initial sentence, he will remain on parole his entire life, dying before he ever begins serving his second sentence.
We believe construing Art. 42.08(b) in this manner leads to an absurd result that the Legislature could not possibly have intended. The goal of the statute is to punish inmates who commit crimes while in prison.10 To require a person, after a long and successful completion of a parole term, to return to prison is less a punishment than it is an absurdity. It seems clear to us that the way to “punish” an inmate who commits an offense while in custody is to delay his release to parole, not require him to return to prison on completion of his parole. The case of the inmate who serves his parole term until the day he dies further heightens the absurdity. That inmate would never be punished at all by the second sentence, as he would never even begin serving it.
The other possible construction of Art. 42.08(b), if “completion of the sentence” means “discharge in full,” is that the inmate not be permitted to make parole at all on his first sentence. In effect, by virtue of having received a second, cumu-lated sentence, he is simply designated as no longer eligible for parole. In this situation, he would have to serve his first sentence day-for-day in prison until it discharged, at which point he could begin his second sentence.
This construction is not facially absurd.11 But it does render meaningless other statutes, and we strive to construe statutes so that all are given effect.12 In particular, § 508.150(a)13 of the Government Code requires the Board to designate a hypothetical parole release date for an inmate serving consecutive sentences under Art. 42.08. This requirement indicates that the Legislature anticipated that an inmate serving stacked sentences would make parole on his first sentence. The requirement does not distinguish between inmates serving sentences stacked pursuant to Art. 42.08(a) versus Art. 42.08(b).
Elimination of parole for a person serving stacked sentences also conflicts with Government Code § 508.145,14 in which the [268]*268Legislature authorized parole in all sentences other than those in which the death penalty is assessed. There is no exception in that statute for inmates serving consecutive sentences. Rather, the statute provides that any sentence other than death is eligible for parole. Finally, elimination of parole creates the potential for “life without parole,” a sentence which has never been legislated in Texas. It is not our duty, as a Court, to create new sentencing provisions or to alter the sentencing scheme which our Legislature has provided.15
So interpreting “completion of the sentence” to mean serving the sentence in full, day-for-day, results in two possible consequences, one of which is absurd, and the other of which puts the statute in conflict with other laws. We must conclude the phrase “completion of the sentence” in Art. 42.08(b) is ambiguous.
C. Extratextual Factors
To resolve the ambiguity, we consult extratextual factors.16 The Legislature has provided a non-exclusive list of factors a court may use to construe a statute, including the: (1) object sought to be obtained; (2) circumstances under which the statute was enacted; (3) legislative history; (4) common law or former statutory provisions, including laws on the same or similar subjects; (5) consequences of a particular construction; (6) administrative construction of the statute; (7) title (caption), preamble, and emergency provision.17
1. Legislative History
We previously considered the legislative history of Art. 42.08(b) in Basden v. State18 and Cannady v. State.19 In those cases, we recognized that the intent of the statute was to deter inmates from committing crimes during their incarceration and to more harshly punish those inmates who are not deterred.20 We have again reviewed the legislative history, and we find no suggestion during the committee hearings or any floor debate that the statute was intended to alter the manner in which TDCJ or the Board administers cumulative sentences.21 Rather, the purpose of the bill was merely to mandate consecutive sentences for inmate offenses, and remove discretion from the trial court to assess a concurrent sentence.22 Perceiving no intent on the Legislature’s part to alter the parole laws, we turn to some of the consequences of different constructions.
2. Consequences of a Particular Construction
We have already considered the consequences of interpreting “completion of sentence” to mean serving the sentence in full, day-for-day. Depending on whether the inmate is permitted to make parole on the first sentence, the consequences of this construction either lead to an absurd result or lead to a conflict with other statutes.
We next consider the consequences of construing “completion of the sentence” to mean something less than discharging the sentence in its entirety. “Completion of the sentence” could mean the point at which the inmate makes parole on the sentence. Under this construction, when the inmate makes parole on the first sentence, he would begin serving his second [269]*269sentence. Although this construction does not coincide with the dictionary definition of “complete,” it is nevertheless a possible interpretation of that phrase. And, notably, it does not lead to an absurd result. Having an inmate make parole on his first sentence and then begin serving his second sentence is certainly administratively feasible. In addition, this construction would not bring Art. 42.08(b) into conflict with any other laws. We discern no adverse consequences to this construction.
3. Laws on Same or Similar Subject
Another aid of statutory construction is to consider laws on the same or similar subjects.23 We consider other uses of the phrase “completion of the sentence.” We also consider the phrase “cease to operate” in Art. 42.08(a).
a. “Completion of the Sentence”
The phrase “completion of the sentence” appears in various other statutes. The Texas Code of Military Justice, located in the Government Code, provides that if a person subject to that act is accused of a criminal offense, he must answer to civil authorities.24 It adds that if that person had previously been court-martialed, he must, after answering to civil authorities, return to military custody “for the completion of the sentence.”25 We have found no caselaw interpreting the phrase or deciding what constitutes “completion” of a sentence of court-martial. But because there is no provision in that Code for early release on a sentence of court-martial, the phrase must mean discharge-in-full. For that same reason (there being no provision for “paroling” a court-martial), the use of the phrase “completion of the sentence” to mean discharge-in-full in this context is not particularly helpful to our case.
Also in the Government Code is a statute providing for deportation of illegal aliens incarcerated in Texas prisons. The statute states that the inmates should be deported “on completion of the inmates’ sentences or release of the inmates on parole or mandatory supervision.”26 In this instance, we can easily determine that “completion of the sentence” refers to discharge in full, because it is listed in a series with other options, including parole and mandatory supervision. The placement of the phrase in a series with parole as another alternative is clear indication that “completion of the sentence” in this context does not contemplate release to parole.
In the Code of Criminal Procedure, Art. 48.05 provides that a person convicted of a federal offense can apply to the Governor for restoration of his civil rights. The statute specifies that the person may not apply until he has “completed the sentence for the federal offense.” There is no definition of “completing the sentence” and no indication of whether making parole constitutes “completing the sentence.” But a common sense interpretation of the statute leads us to believe that the Governor would not restore a person’s civil rights until he has discharged the sentence in full; simply making parole is not enough.
Article 56.11(a)(1) provides for notifying victims of the release or escape of an inmate. The statute states that the victim should be notified when the person convicted of the offense “completes the person’s sentence and is released.”27 In this statute, “completing the sentence” appears to include both discharge-in-full and release to parole. The statute’s purpose is to inform victims when an individual has been released to society. Common sense tells us that release from prison occurs through parole more often than through discharge. So for this statute to have any [270]*270meaning at all, “completion of the sentence” must include release to parole.
Although most uses of the phrase “completion of the sentence” in our statutes appear to refer to discharge of the sentence in full, one statute’s use of the phrase seems to include either discharge or parole. So interpreting “completion of the sentence” in Art. 42.08(b) to include release to parole would not be unprecedented.
b. “Cease to Operate”
We next consider the meaning of the phrase “cease to operate” in Art. 42.08(a). As noted previously, subsection (a) refers to the first sentence “ceasing to operate,” while subsection (b) uses the phrase “completion of the sentence.” The use of different language in the two subsections could suggest that different meanings were intended. On the other hand, it may simply have been the result of the Legislature differentiating between cases in which a trial court must, as opposed to may, order a cumulative sentence. In any event, defining “cease to operate” may assist us in defining “completion of the sentence.”
The phrase “cease to operate” is specifically defined in the provisions governing the applicability of parole to consecutive sentences. The current definition is essentially the same as it was at the time of Kuester’s offense:
(2) For the purposes of Article 42.08, Code of Criminal Procedure, the judgment and sentence of a prisoner sentenced for a felony, other than the last sentence in a series of consecutive sentences, cease to operate:
(A) when the actual calendar time served by the inmate equals the sentence imposed by the court; or
(B) on the date a parole panel designates as the date the inmate would have been eligible for release on parole if the inmate had been sentenced to serve a single sentence.28
The meaning of subsection (A) is clear enough: when the inmate has served the first sentence in full, day-for-day, he can begin serving his second sentence. The text of subsection (B) is less clear. What constitutes “the date a parole panel designates as the date on which the prisoner would have been eligible for release on parole”? Is this the date that the person becomes statutorily eligible for parole? Or is it the date, after becoming statutorily eligible, that the prisoner would have actually been released to parole upon approval of the Board?
The Board has the duty to determine when an offender may be released from prison to parole supervision.29 In parole matters, the Board acts in panels composed of three members.30 Since the statute refers to a parole panel’s designation, rather than an inmate’s statutory eligibility, we are swayed that subsection (B) refers to a discretionary decision on the part of the Board that the person actually would have been released to parole but for the second sentence. It is not enough that the inmate merely becomes statutorily eligible for parole. In order for his sentence to cease to operate, a parole panel must designate the date that the inmate was approved for parole and would have been released to parole but for the second consecutive sentence. So we conclude that “cease to operate” means the date the prisoner has served the sentence day-for-day in TDCJ-ID, or the date a parole panel has designated that the prisoner would have been released on parole if he had been sentenced to serve only one sentence.
4. Placement and Language of Art. 42.08
We next consider the placement of Art. 42.08 in the Code of Criminal Procedure. [271]*271Notably, all of the laws governing parole were originally found in Art. 42.18 of that Code and can now be found in the Government Code. The separate location of Art. 42.08 indicates that the Legislature did not view the statute as affecting parole law. In addition, Art. 42.08’s language is directed to the trial court, not to custodial officials. The statute’s placement and its language are both some support for the view that the first sentence should not have to be served in full, day-for-day, in order for the second sentence to begin. Rather, the parole laws should apply to stacked sentences as they do to all other sentences, and the second sentence should begin when the inmate makes parole on the first sentence.
D. Meaning of “Completion of Sentence”
Considering the consequences of varying constructions, the legislative history, the meaning of “completion of the sentence” in other statutes, the meaning of “cease to operate” in Art. 42.08(a), the language of Art. 42.08(b), and its placement in the Code, we conclude that the phrase “completion of the sentence” means serving the sentence in full, day-for-day, until discharge, or receiving a vote of approval for parole by a parole panel. Granted, this results in “completion of the sentence” having the same meaning as “cease to operate.” But we are persuaded that the use of different phrases in subsections (a) and (b) of Art. 42.08 is insignificant compared to the other considerations in this opinion. So a sentence is “completed,” for purposes of Art. 42.08(b), when it is served out in full day-for-day, or when a parole panel approves the inmate for parole release.
III. Erroneous Release
With this definition in mind, we address Kuester’s claim that he was erroneously released to parole in 1992. He argues he should not have been released to parole in 1992, but instead should have begun serving his second sentence. He seeks credit on his assault sentence for the period of his erroneous release.31
On September 8, 1992, Kuester was granted parole. At that time, TDCJ-ID had improperly computed Kuester’s parole eligibility on the fourteen year total of his cumulative sentences. Nevertheless, despite this error, Kuester was also eligible for parole on his ten-year conviction had it been calculated singularly. Under the law in effect on the date of his offense, Kuester became eligible for parole on that sentence after his accrued calendar time and good time conduct exceeded one-fourth of his sentence.32 One-fourth of Kuester’s 10-year sentence is 2½ years. At the time Kuester was released to parole, he had accrued more than four years of actual calendar time. So Kuester’s release to parole on his burglary offense was permissible, even though the decision was made using an erroneous time calculation. As a result, Kuester’s initial sentence ceased to operate on that date, and he should have begun serving his aggravated assault conviction at that time.
Upon starting his second sentence, Kuester should have been given credit for the time he served in jail before trial.33 His judgment granted him 552 days of pretrial credit. Kuester was eligible for release to parole on his aggravated assault conviction after he had served one-quarter of his sentence, considering both good and flat time.34 One-fourth of Kuester’s four-year sentence is one year. Giving him credit for the 552 days, Kuester was eligi[272]*272ble for parole on this sentence the same day that he began serving it.
So despite the fact that Kuester’s sentences had been improperly collectively calculated, on September 8, 1992, Kuester was in fact parole eligible on both his first and second sentence. His claim that he was erroneously released is not supported by the facts. Since his release was not erroneous, he is not entitled to credit for the time that he spent on parole.
IV. Credit after revocation
Although Kuester is not entitled to credit for time he spent on parole, we still must consider whether he is entitled to credit from the date his parole was revoked. The issue we face is what happens when an inmate is paroled on two consecutive sentences. Is he serving his parole on both sentences concurrently, even though the sentences were cumulated? If so, what happens when his parole gets revoked? Does he continue to serve his sentences concurrently, or do the sentences return to their consecutive status when the inmate returns to prison?
One option would be for consecutive sentences to be, in essence, “re-stacked” upon release to parole and then “re-stacked” again upon parole revocation. That is, the inmate serves the first sentence until he is approved for parole, then serves the second sentence until he makes parole. Upon his release to parole, he goes back to serving the first sentence on parole, serving only that sentence until it discharges, at which point he begins serving the second sentence on parole. If his parole gets revoked while serving the first sentence, he returns to custody only on that first sentence, and serves it alone until he is approved for parole. At that point, he begins serving his second sentence again. And so on.
The problem with this scenario is our well-settled law that once a person begins serving a sentence, that person continues serving the sentence, either in prison, on parole, or on mandatory supervision, until the sentence is discharged. A sentence must be continuous and a prisoner or inmate cannot be required to serve his sentence in installments.35
Instead, we return to the definition of “cease to operate” in § 508.150(b)(2) of the Government Code for assistance. In subsection (A), a sentence can cease to operate by being fully served, day-for-day. When a sentence ceases to operate under this subsection, the next sentence in the series of consecutive sentences begins. A cumulated sentence in that situation does not begin to operate until the previous sentence is discharged. Under this scheme, the effective result is literally “stacked” sentences.
In contrast, subsection (B) is not a “stacking” of sentences in the literal sense. Under subsection (B), a sentence other than the last in a series of consecutive sentences ceases to operate when a parole panel designates the date on which the prisoner would have been eligible for release on parole if the prisoner had been sentenced to serve a single sentence. When a sentence ceases to operate under subsection (B), the next sentence in the series of consecutive sentences begins. The difference between the operation of cumulative sentences under subsection (A) and subsection (B) is that under subsection (B) the previous sentence has not yet been discharged when the next sentence begins.
Under subsection (B), when the first sentence ceases to operate and the second sentence begins, the inmate continues serving the first sentence while he serves the second sentence. That is, the first sentence ceases to operate for purposes of Art. 42.08, but it does not stop altogether. While the inmate is in custody serving his [273]*273second sentence, he is “paroled” on his first sentence, continuing to get credit on it for the time he is in prison. Upon making parole on the second sentence, the inmate serves both sentences concurrently, both while he is on parole and once his parole is revoked. The first sentence ceases to operate only for the purpose of determining when the second sentence begins. It does not literally cease to operate.
Application
On September 8, 1992, the Board released Kuester on parole. At that point Kuester was on parole for both sentences until his parole was revoked in 1993. After Kuester’s parole was revoked for both sentences and he was returned to prison, he was serving both sentences concurrently since he had already begun serving each sentence.36 (We note that denying Kuester credit for the time he spent on parole does not mean the sentences were being served in installments. The statute provides for the loss of credit for the parole period following revocation of parole.37) Kuester is entitled to credit for time served on his four-year sentence from August 29, 1993, which was the date of his arrest on the parole violator warrant, to the present.
So Kuester is entitled to the following credit on his four-year sentence: 552 days of pretrial credit shown in the judgment; credit for one day on September 8, 1992, which was the date he began serving his four-year sentence and the date on which he was released to parole; and credit from August 29, 1993, to the present, which was the date he was arrested on the parole violator warrant leading to the revocation of parole for both sentences. The record indicates that Kuester has not received all of that credit on his four-year sentence and that, upon receiving that credit, the sentence will be fully served.
V. Judgment
Relief is granted. The Texas Department of Criminal Justice-Institutional Division is ordered to amend Kuester’s records to reflect the credit in Cause Number 20,657, in the 3rd District Court of Anderson County, as follows: 552 days of pretrial credit shown in the judgment; credit for one day on September 8, 1992; and credit from August 29, 1993, to the present. We further order Kuester’s immediate release from further confinement in Cause Number 20,657.
Copies of this opinion shall be sent to the Texas Department of Criminal Justice, Institutional Division, Paroles Division, and Board of Pardons and Paroles Division.
KELLER, J., delivered a dissenting opinion.