【正文】 eject the traditional assumption that a single translation method can be appropriate for all circumstances in which translations occur and for all purposes that translation serves. Circumstances underlying foreign exchange translation differ widely. Translating accounts from a stable to an unstable currency is not the same as translating accounts from an unstable currency to a stable one. Likewise, there is little similarity between translations involving importor exporttype translations and those involving a permanently established affiliate or subsidiary pany in another country that reinvests its local earnings and does not intend to repatriate any funds to the parent pany in the near future. Second, translations are made for different purposes. Translating the accounts of a foreign subsidiary to consolidate those accounts with those of the parent pany has very little in mon with translating the accounts of an independent pany mainly for the convenience of various foreign audiencesof –interest. We therefore pose three questions: 1. Is it reasonable to use more than one translation method? 2. If so, what should be the acceptable methods and under what conditions should they be applied? 3. Are there situations in which translations should not be done at all? As to the first question, it is clear that a single translation method cannot equally serve translations occurring under different conditions and for different purposes. More than one translation method is needed. Regarding the second question, we think that three different translation approaches can be accepted: (1) the historical method, (2) the current method, and (3) no translation at all. Financial accounts of foreign entities can be translated either from a parent pany perspective or from a local perspective. Under the parent pany perspective, foreign operations are extensions of parent pany operations and are, in large measure, sources of domestic currency cash flows. Accordingly, the object of translation is to change the unit of measure for financial statements of foreign subsidiaries to the domestic currency, and to make the foreign statements conform to accounting principles generally accepted in the country of the parent pany. We think these objectives are best achieved by translation methods that use historical rates of exchange. We prefer the temporal principle, as it generally maintains the accounting principles used to measure assets and liabilities originally expressed in foreign currency units. Because foreign statements under a parent pany perspective are first adjusted to reflect parent pany accounting principles (before translation), the temporal principle is appropriate, as it changes a measurement in foreign currency into a measurement in domestic currency without changing the basis of measurement. The temporal translation method is easily adapted to processes that make accounting adjustments during the translation. When this is so, adjustments for differences between two or more sets of accounting concepts and practices are made along with the translation of currency amounts. For example, inventories or certain liabilities may be restated according to accounting practices different from those originally used. The temporal principle can acmodate any asset valuation framework, be it historical cost, current replacement price, or realizable values. The current rate method of translation is a straightforward translation (restatement) from one currency language to another. There is no change in the nature of the accounts。 that is, when foreign entities are viewed from a local (as opposed to a parent) pany perspective. Translation at the current rate does not change any of the initial relationships (., financial ratios) in the foreign currency statements, as all account balances are simply multiplied by a constant. This approach is also useful when the accounts of an independent pany are translated for the convenience of foreign stockholders or other external user groups. A second use of the current rate method happens when priceleveladjusted accounts are to be translated to another currency. If reliable price level adjustments are made in a given set of accounts and if domestic price level changes for the currency are reflected closely in related foreign exchange rate movements, the current rate translation of priceleveladjusted data yields results that are parable to translating historical cost accounts under the historical rate translation method. Are there situations in which translations should not be done at all? We think so. No translation is appropriate between highly unstable and highly stable currencies. Translation of one into the other will not produce meaningful information using any translation method. No translation also means nonconsolidation of financial statements. We think this is reasonable. If a currency is unstable enough to put account translations out of the question, financial statement