This case involves the interpretation of an insurance policy. Appellant James E. Elam *293purchased a long-term-disability policy from Appellee First Unum Life Insurance Company, through his employer, Smith-Barney. In 1994, Elam elected coverage under the policy for disability due to his having bipolar affective disorder. First Unum recognized Elam’s disability and paid benefits to him from August 9, 1994, through August 9, 1996. First Unum declined to pay further benefits because of a twenty-four-month policy limitation applicable to disability due to mental illness. Elam filed suit against First Unum, contending that the term “mental illness,” as used and defined in the policy, is ambiguous as it applies to his illness. The crux of Elam’s argument is that bipolar affective disorder is a physical illness, with a biological origin, that is treated with medication. The Garland County Circuit Court granted summary judgment to First Unum and found that the plain, ordinary meaning of the term encompassed bipolar affective disorder. We reverse the trial court’s judgment and remand for trial.
The insurance policy at issue provides: “Benefits for disability due to mental illness will not exceed 24 months of monthly benefit payments,” unless the insured meets one of two possible situations not applicable in this case. The policy then provides: “ ‘Mental illness’ means mental, nervous or emotional diseases or disorders of any type.” The parties do not dispute the fact that Elam suffers from bipolar affective disorder. Instead, they dispute whether bipolar affective disorder is a “mental illness” within the meaning of that term as set out in the disability policy.
The record demonstrates that Elam filed a motion for summary judgment, asserting that there were no material issues of fact to be resolved and that the court need only determine the legal issue of whether the policy language was ambiguous. Elam contended that the affidavits of Dr. Bradley C. Diner, M.D., and Dr. Charles L. Bowden, M.D., showed that bipolar affective disorder was a physical, rather than mental, illness. Dr. Diner stated in his affidavit that “[i]t is well accepted in the scientific community that bipolar disorder is a biological condition with hereditary predisposition, and an alteration in brain chemistry is being responsible for the mood disturbance and altered thought process.” Similarly, Dr. Bowden stated that, in his opinion, bipolar affective disorder has a biological origin. Dr. Bowden also opined that “there is no longer any reasonable doubt among informed members of the medical community that Bipolar Affective Disorder has a biological origin” and that this principle has gained “near universal acceptance in the medical community.”
*294First Unum also filed a motion for summary judgment, contending that the policy plainly and unambiguously excluded further benefits to Elam for disability due to mental illness and that bipolar affective disorder is a mental illness. In support of its motion, First Unum submitted portions of depositions given by Dr. Diner and Dr. Joe Backus, M.D., Elam’s treating physician. In particular, First Unum relied on the following information: (1) the precise cause of bipolar affective disorder is unknown, as it cannot be detected by such diagnostic tools as brain scans or blood tests; (2) diagnosis is based primarily on the patient’s behavior, clinical presentation, and history; (3) psychiatrists are unable to predict the onset of symptoms based on any physical or biological indicator; (4) individuals suffering from the disorder are treated exclusively by psychiatrists, using a combination of drugs and psychotherapy; and (5) the medical and psychiatric communities officially classify bipolar affective disorder as a mental illness, regardless of their opinion that the disorder has a biological origin.
In response to First Unum’s motion, Elam submitted the depositions of Drs. Diner and Backus in their entirety, which reflected the following information: (1) bipolar affective disorder is caused by an imbalance of chemicals or neurotransmitters in the brain; (2) the disorder is distinguishable from a personality disorder, which is triggered by social stress, such as the death of a loved one; and (3) the disorder is treated with medication, to stabilize the neurotransmitter blood levels in the brain, while psychotherapy is used primarily to help educate the patient about the illness and to help him adapt to the life changes that the illness brings. Elam also offered his own affidavit, to which were attached approximately twenty-five articles from various sources, including newspapers, magazines, mental-health pamphlets, and a law journal. Collectively, these sources report that certain disorders of the brain, including bipolar affective disorder, are no longer viewed as mental illnesses due to the apparent consensus in the medical community that such disorders have physical origins. Finally, Elam relied on the affidavit of Dr. Diner, which reflected that there is a general trend in psychiatry to move away from the classification of illnesses by symptoms in favor of classifying them according to their causes or origins. As an example of this trend, Dr. Diner’s affidavit contained the following disclaimer from the introduction to the Fourth Edition of the Diagnostic and Statistical Manual (DSM-IV):
*295Although this volume is titled The Diagnostic and Statistical] Manual of Mental Disorders, the term mental disorder unfortunately implies a distinction between “mental” disorders and “physical” disorders that is a reductionistic anachronism of mind/body dualism. A compelling literature documents that there is much “physical” in “mental” disorders and much “mental” in “physical” disorders. The problem raised by the term “mental” disorders has been much clearer than its solution, and, unfortunately, the term persists in the title of the DSM-IV because we have not found an appropriate substitute.
After considering the evidence offered by both parties, the trial court ruled that the definition of “mental illness” provided in the disability policy is unambiguous when the term is afforded its plain and ordinary meaning. The trial court , found that Arkansas law requires that contract terms, such as those contained in insurance policies, be interpreted according to a layperson’s understanding of those terms. The trial court concluded:
The common, ordinary and lay understanding of the term “mental illness” encompasses bi-polar affective disorder. This interpretation is further supported by the testimony of Plaintiff s treating and consulting psychiatrists, Dr. Bradley Diner and Dr. Joe Backus, that the psychiatric community classifies bi-polar affective disorder as a mental illness regardless of the precise cause or etiology of that disease. As used in Defendant’s insurance policy, that term is not ambiguous.
Based on these conclusions, the trial court granted summary judgment to First Unum.
Elam appealed to the Arkansas Court of Appeals. After reviewing the evidence submitted by the parties, as weH as varying case law on this issue from other jurisdictions, the court of appeals concluded that the policy’s language was ambiguous. It thus reversed the trial court’s judgment and remanded for trial on the issue of whether Elam’s disorder is a mental illness. See Elam v. First Unum Life Ins. Co., 72 Ark. App. 54, 32 S.W.3d 486 (2000). Both parties filed for review of that decision. Elam sought review on the ground that the court of appeals erred when it refused to enter judgment in his favor once it concluded that the policy was ambiguous. First Unum, on the other hand, sought review of the initial determination that the policy was ambiguous. We granted review pursuant to Ark. Sup. Ct. R. 1-2(e). When we grant review following a decision by the court of appeals, we review the case as though *296it had been originally filed in this court. See Regions Bank & Trust v. Stone County Skilled Nursing Facil., Inc., 345 Ark. 555, 49 S.W.3d 107 (2001); Freeman v. Con-Agra Frozen Foods, 344 Ark. 296, 40 S.W.3d 760 (2001). Thus, we review the trial court’s judgment, not that of the court of appeals.
Elam argues that the trial court erred in granting summary judgment to First Unum and concluding, as a matter of law, that the “mental illness” limitation in the insurance policy is unambiguous. We agree that summary judgment was improper, based on the circumstances of this case.
Summary judgment should only be granted when it is clear that there are no genuine issues of material fact to be litigated, and the moving party is entided to judgment as a matter of law. City of Lowell v. City of Rogers, 345 Ark. 33, 43 S.W.3d 742 (2001); Elam v. Hartford Fire Ins. Co., 344 Ark. 555, 42 S.W.3d 443 (2001). The purpose of summary judgment is not to try the issues, but to determine whether there are any issues to be tried. BPS, Inc. v. Parker, 345 Ark. 381, 47 S.W.3d 858 (2001); Flentje v. First Nat’l Bank of Wynne, 340 Ark. 563, 11 S.W.3d 531 (2000). Thus, summary judgment is not proper, where, although the actual facts are not in dispute, they may result in differing conclusions as to whether the moving party is entitled to judgment as a matter of law. Wallace v. Broyles, 332 Ark. 189, 961 S.W.2d 712 (1998) (per curiam) (citing Ark. R. Civ. E 56(c); Thomas v. Sessions, 307 Ark. 203, 818 S.W.2d 940 (1991)). See also City of Lowell, 345 Ark. 33, 43 S.W.3d 742; Ultracuts Ltd. v. Wal-Mart Stores, Inc., 343 Ark. 224, 33 S.W.3d 128 (2000). In other words, where different conclusions may be reached regarding what the undisputed facts demonstrate, summary judgment is inappropriate.
The parties do not dispute the underlying facts of this case, i.e., that Elam suffers from bipolar affective disorder and has suffered from the illness for years. What is disputed is the meaning of the term “mental illness,” as it is used and defined in the insurance policy issued to Elam by First Unum. The parties essentially agreed that resolution of this issue required the trial court to look beyond the face of the policy, as both parties submitted disputed extrinsic evidence in an attempt to define the term. The trial court’s order demonstrates that it relied on that extrinsic evidence in determining that the term was commonly understood and thus unambiguous. This was error.
*297 The law regarding construction of an insurance contract is well settled. If the language of the policy is unambiguous, we will give effect to the plain language of the policy without resorting to the rules of construction. Norris v. State Farm Fire & Cas. Co., 341 Ark. 360, 16 S.W.3d 242 (2000); Western World Ins. Co. v. Branch, 332 Ark. 427, 965 S.W.2d 760 (1998). On the other hand, if the language is ambiguous, we will construe the policy liberally in favor of the insured and strictly against the insurer. Id. Language is ambiguous if there is doubt or uncertainty as to its meaning and it is fairly susceptible to more than one reasonable interpretation. Norris, 341 Ark. 360, 16 S.W.3d 242; Smith v. Prudential Prop. & Cas. Ins. Co., 340 Ark. 335, 10 S.W.3d 846 (2000). Ordinarily, the question of whether the language of an insurance policy is ambiguous is one of law to be resolved by the court. Norris, 341 Ark. 360, 16 S.W.3d 242; Western World, 332 Ark. 427, 965 S.W.2d 760. Where, however, parol evidence has been admitted to explain the meaning of the language, the determination becomes one of fact for the jury to determine. See Smith, 340 Ark. 335, 10 S.W.3d 846; Southhall v. Farm Bureau Mut. Ins. Co., 276 Ark. 58, 632 S.W.2d 420 (1982).
Our case law demonstrates that where there is a dispute as to the meaning of a contract term or provision, be it an insurance or other contract, the trial court must initially perform the role of gatekeeper, determining first whether the dispute may be resolved by looking solely to the contract or whether the parties rely on disputed extrinsic evidence to support their proposed interpretation. As Justice George Rose .Smith explained, “[t]he construction and legal effect of written contracts are matters to be determined by the court, not by the jury, except when the meaning of the language depends upon disputed extrinsic evidence.” Id. at 60, 632 S.W.2d at 421 (emphasis added). Thus, where the issue of ambiguity may be resolved by reviewing the language of the contract itself, it is the trial court’s duty to make such a determination as a matter of law. On the other hand, where the parties go beyond the contract and submit disputed extrinsic evidence to support their proffered definitions of the term, this is a question of fact for the jury. In the latter situation, summary judgment is not proper.
We cannot say that the term “mental illness” is ambiguous looking only to the face of the policy. Rather, the only way that term may become ambiguous is by looking to the conflicting evidence submitted by the parties, which included the affidavits and depositions of three physicians, an excerpt from the DSM-IV, and articles from various sources. Both parties went beyond the four *298corners of the policy and relied on this extrinsic evidence to support their respective interpretations of the term “mental illness.” Elam claimed that the evidence showed that the term did not include those illnesses having physical causes, while First Unum asserted that, regardless of the cause, the evidence showed that bipolar affective disorder is classified, diagnosed, and treated as a mental illness by psychiatrists. Accordingly, the trial court erred in resolving this issue as a matter of law by granting summary judgment to First Unum. We therefore reverse the trial court’s judgment and remand for the jury to resolve whether, based on the disputed extrinsic evidence offered by the parties, the term “mental illness,” as it is used and defined in the policy, is ambiguous as it applies to Elam’s diagnosis of bipolar affective disorder. In resolving this issue, the jury shall be instructed on the relevant law regarding construction of an insurance contract.
Circuit Court reversed and remanded; Court of Appeals reversed.
GLAZE, J., not participating.