By Caleb Younger1

The Sixth Amend­ment states that “in all crim­i­nal pros­e­cu­tions, the accused shall enjoy the right . . . to be con­front­ed with the wit­ness­es against him.”2 The right to con­front and cross-exam­ine wit­ness­es who pro­duce tes­ti­mo­ny is there­fore guar­an­teed by the Con­sti­tu­tion. Pri­or to 2004, courts deter­mined the admis­si­bil­i­ty of hearsay state­ments at least par­tial­ly based on reli­a­bil­i­ty con­cerns.3 The lan­guage con­duit the­o­ry pre­vailed when deter­min­ing the admis­si­bil­i­ty of state­ments made by inter­preters.4 Courts deter­mined whether the inter­preter was a con­duit – a mere chan­nel through which the state­ment pas­sive­ly changed from one lan­guage to anoth­er – of the orig­i­nal speak­er by using fac­tors that looked at whether an inter­preter dur­ing police inter­views pro­vid­ed an accu­rate and reli­able trans­la­tion; if they did, then inter­pre­ta­tions could be used as evi­dence with­out the inter­preter being avail­able for cross-exam­i­na­tion.5

When the Supreme Court rein­ter­pret­ed the Con­fronta­tion Clause in Craw­ford v. Wash­ing­ton, the major­i­ty opin­ion reject­ed this bal­anc­ing of fac­tors in favor of a clear-cut rule pro­hibit­ing the admis­sion of tes­ti­mo­ni­al evi­dence in the absence of an oppor­tu­ni­ty for con­fronta­tion.6 Lat­er Supreme Court cas­es fur­ther devel­oped the require­ments for the Con­fronta­tion Clause. In Bull­com­ing v. New Mex­i­co, the Court held that foren­sics tests could not be admit­ted as evi­dence unless the lab­o­ra­to­ry ana­lyst who did the test was made avail­able for cross-exam­i­na­tion.7

But the future of the lan­guage con­duit the­o­ry remains uncer­tain. The Fourth, Sev­enth, Eighth, and Ninth Cir­cuits have all found that the the­o­ry still applies to inter­preters post-Craw­ford.8  These courts’ opin­ions, how­ev­er, raise ques­tions about how Craw­ford’s uncom­pro­mis­ing stan­dard fits with the extant legal frame­work with regards to the admis­si­bil­i­ty of inter­preters’ state­ments. The Eleventh Cir­cuit, on the oth­er hand, has held that Craw­ford requires a retreat from the lan­guage con­duit the­o­ry.9

This Con­tri­bu­tion will ulti­mate­ly argue that the Craw­ford Court prop­er­ly inter­pret­ed the Sixth Amend­ment and that the lan­guage con­duit the­o­ry fails under both Supreme Court jurispru­dence and the Con­sti­tu­tion­al text. Inter­preters are new declar­ants pro­vid­ing tes­ti­mo­ni­al wit­ness against a defen­dant, and thus must be avail­able under the Con­fronta­tion Clause.

* * * * *

In 1912, the Sev­enth Cir­cuit first ruled that when a per­son whose dec­la­ra­tion is admis­si­ble “assent[s] to the use of an inter­preter… they nec­es­sar­i­ly assume that the inter­preter is trust­wor­thy, which makes his lan­guage pre­sump­tive­ly their own.”10 Although the court did not use the words “lan­guage con­duit” in this opin­ion, lat­er cas­es accept­ed and expand­ed on this the­o­ry.11 These lan­guage con­duit the­o­ries devel­oped in the shad­ow of Ohio v. Roberts, where­in the Supreme Court prece­dent affirmed that “indi­cia of reli­a­bil­i­ty” were cen­tral to deter­min­ing whether hearsay state­ments could be admit­ted when the declar­ant was absent.12

In Craw­ford v. Wash­ing­ton, the Court set out a new frame­work to deter­mine whether hearsay state­ments could be admit­ted, over­turn­ing the prece­dent of Ohio v. Roberts.13 The Court found that “[a]dmitting state­ments deemed reli­able by a judge is fun­da­men­tal­ly at odds with the right of con­fronta­tion.“14 Instead, “[w]here tes­ti­mo­ni­al state­ments are at issue, the only indi­ci­um of reli­a­bil­i­ty suf­fi­cient to sat­is­fy con­sti­tu­tion­al demands is the one the Unit­ed States Con­sti­tu­tion actu­al­ly pre­scribes: con­fronta­tion.”15 If a state­ment is tes­ti­mo­ni­al, then it impli­cates the Con­fronta­tion Clause of the Sixth Amend­ment.16 The only way that inter­preters will not fall under the Con­fronta­tion Clause is if they are (1) not pro­vid­ing tes­ti­mo­ny and (2) are a sep­a­rate witness.

Tes­ti­mo­ni­al state­ments include “mate­r­i­al such as affi­davits. . . or sim­i­lar pre­tri­al state­ments that declar­ants would rea­son­ably expect to be used pros­e­cu­to­ri­al­ly.”17 To be con­sid­ered tes­ti­mo­ni­al, a state­ment must have the pri­ma­ry pur­pose of estab­lish­ing past events poten­tial­ly rel­e­vant to crim­i­nal pros­e­cu­tion.18 And “state­ments tak­en by police offi­cers in the course of inter­ro­ga­tions are defin­i­tive­ly tes­ti­mo­ni­al,” as are wit­ness state­ments giv­en to inves­ti­gat­ing offi­cers.19  The Con­fronta­tion Clause does not dis­tin­guish between accu­rate and inac­cu­rate tes­ti­mo­ny, requir­ing only that any tes­ti­mo­ny be sub­ject to cross-examination.

This log­ic explains the hold­ing in Bull­com­ing v. New Mex­i­co that foren­sics reports could not be admit­ted with­out allow­ing con­fronta­tion of the per­son who pre­pared the report.20  The Supreme Court wasn’t sole­ly con­cerned with reli­a­bil­i­ty and degrees of accu­ra­cy, hold­ing that the prop­er method for that deter­mi­na­tion was thor­ough cross-exam­i­na­tion by the defen­dant.21 Unless the defen­dant had the abil­i­ty to cross-exam­ine the declar­ant of a report, even one as bina­ry as whether a sub­stance was cocaine or not, then it was inad­mis­si­ble.22

The Sixth Amendment’s Con­fronta­tion Clause can­not do any work if the per­son pro­vid­ing tes­ti­mo­ni­al evi­dence is the defen­dant the tes­ti­mo­ny is being used against – one can­not claim the right to con­fronta­tion against one­self. The lan­guage con­duit the­o­ry main­tains that the inter­preter is mere­ly a con­duit whose Eng­lish trans­la­tion is prop­er­ly imput­ed to the orig­i­nal speak­er and there­fore that the interpreter’s trans­la­tion “trans­la­tion is attrib­ut­able to the defen­dant as his own admis­sion and is prop­er­ly char­ac­ter­i­z­able as non-hearsay.”23

This rea­son­ing that an interpreter’s trans­la­tions are the defendant’s dec­la­ra­tions was not con­sid­ered absolute, how­ev­er.  The Ninth Cir­cuit pro­vid­ed a list of fac­tors which have been used by mul­ti­ple courts since.24  These fac­tors looked at (1) which par­ty sup­plied the inter­preter; (2) whether the inter­preter had any motive to mis­lead; (3) the interpreter’s skill and qual­i­fi­ca­tion in the lan­guage; and (4) whether sub­se­quent actions by the defen­dant were con­sis­tent with the trans­la­tion.25 This fact-spe­cif­ic, case-by-case inquiry could look at both the reli­a­bil­i­ty of the inter­preter and whether they were actu­al­ly serv­ing as a con­duit for the defendant.

After Craw­ford, a num­ber of cir­cuits, includ­ing the Fourth, Sev­enth, Eighth, and Ninth, reaf­firmed their prece­dents.26 In Unit­ed States v. Orm Hieng, the Ninth Cir­cuit stat­ed that “we can apply Nazemi­an with­out run­ning afoul of Craw­ford.”27 That case rea­soned that Craw­ford analy­sis looked at hearsay state­ments, but if the state­ments were attrib­uted to the defen­dant under the lan­guage con­duit the­o­ry, they were non-hearsay.28 The Ninth Cir­cuit did so even after the Supreme Court’s rul­ing in Bull­com­ing v. New Mex­i­co.29  This rea­son­ing did track with past prece­dent, such as the Sec­ond Cir­cuit in Unit­ed States v. Da Sil­va.30  Rather than apply­ing the lan­guage con­duit the­o­ry to sug­gest that an inter­preter was the same declar­ant as the defen­dant, the court held that the defen­dant had explic­it­ly autho­rized the inter­preter to speak on their behalf.31  Because this fit into a non-hearsay excep­tion, it could be admitted.

The Eleventh Cir­cuit took a dif­fer­ent view of the Supreme Court’s new frame­work and its effect on the lan­guage con­duit the­o­ry. It held in Unit­ed States v. Charles that the lan­guage con­duit the­o­ry could no longer sur­vive post-Craw­ford because it looked at reli­a­bil­i­ty of an inter­preter instead of whether the inter­preter act­ed as a tes­ti­mo­ni­al wit­ness or new declar­ant.32  The Eleventh Cir­cuit wrote that “even though an interpreter’s state­ments may be per­ceived as reli­able and thus admis­si­ble under the hearsay rules, the Court, in Craw­ford, reject­ed reli­a­bil­i­ty as too nar­row a test for pro­tect­ing against Con­fronta­tion Clause vio­la­tions.”33 And they also not­ed that “none of the [Nazemi­an] fac­tors… bear upon the basic fact that the inter­preter is the speak­er (declar­ant) of the out-of-court Eng­lish lan­guage state­ments”34 The Eleventh Cir­cuit cre­at­ed this new cir­cuit split because it cor­rect­ly read the Supreme Court’s post-Craw­ford frame­work as requir­ing a deter­mi­na­tion not of inter­preter reli­a­bil­i­ty, but a deter­mi­na­tion of the iden­ti­ty of a tes­ti­mo­ni­al wit­ness.35

* * * * *

The Eleventh Circuit’s view is cor­rect because inter­preters are pro­vid­ing tes­ti­mo­ny as sep­a­rate declar­ants. In any police inter­view, both ini­tial respons­es by the defen­dant and the translator’s inter­pre­ta­tions to an offi­cer are tes­ti­mo­ni­al, as they can both expect that their words estab­lish past events and might be used pros­e­cu­to­ri­al­ly.36 Inter­pre­ta­tion in police inter­views will always be tes­ti­mo­ni­al – the strug­gle is mere­ly over who is actu­al­ly pro­vid­ing this tes­ti­mo­ny.  The lan­guage con­duit the­o­ry fails because it deter­mines admis­si­bil­i­ty on reli­a­bil­i­ty grounds instead of look­ing at whether there are sep­a­rate declar­ants for testimony.

A defendant’s state­ment in one lan­guage and the interpreter’s trans­la­tion into a dif­fer­ent lan­guage are two sep­a­rate dec­la­ra­tions by two sep­a­rate declar­ants. Because inter­pre­ta­tion “does not pro­vide for a one-to-one cor­re­spon­dence between words or con­cepts in dif­fer­ent lan­guages,” the trans­la­tor must make assump­tions and asser­tions at every moment because lan­guages do not cor­re­spond exact­ly.37 A trans­la­tor “ma[kes] rep­re­sen­ta­tions each time he translate[s] state­ments from one lan­guage to anoth­er.”38 While the defen­dant is always the declar­ant of the words they speak, the inter­preter is the declar­ant of the Eng­lish state­ments.39  Because the trans­la­tion is a new expres­sion of ideas—even assum­ing it accu­rate­ly or reli­ably com­mu­ni­cates the ideas of the ini­tial declarant—it is new tes­ti­mo­ni­al evi­dence by a new declar­ant.40

The Nazemi­an fac­tors look to the reli­a­bil­i­ty of an inter­preter but do not ade­quate­ly prove that an inter­preter is not a declar­ant in their role as a con­duit.41 While there are poten­tial cas­es where an inter­preter is an agent of the defen­dant, that is best dealt with through agent autho­riza­tion frame­works and not the lan­guage con­duit theory.

The Supreme Court’s rul­ing in Bull­com­ing v. New Mex­i­co also weighs against the lan­guage con­duit the­o­ry.42 In that case, the Court held that prepar­ing a foren­sics report made the pre­par­er a new declar­ant.43 Inter­preters exer­cise more dis­cre­tion than lab­o­ra­to­ry ana­lysts because “trans­la­tion from one lan­guage to anoth­er is much less of a sci­ence than con­duct­ing lab­o­ra­to­ry tests, and so much more sub­ject to error and dis­pute.”44 That means that the impor­tance of poten­tial con­fronta­tion is high­er and can have greater pro­ba­tive val­ue. The lack of sci­en­tif­ic pre­ci­sion in a one-to-one rela­tion­ship only demon­strates the need for confrontation.

* * * * *

The lan­guage con­duit the­o­ry might still be used by the major­i­ty of cir­cuits, but after Craw­ford and Bull­com­ing, the Eleventh Circuit’s approach is more con­sis­tent with Supreme Court prece­dent.  As inter­preters’ dec­la­ra­tions are sep­a­rate tes­ti­mo­ni­al state­ments, to com­ply with the Con­fronta­tion Clause, they must be avail­able for con­fronta­tion or their out-of-court state­ments should not be admitted.



1. Caleb Younger is a 3L at New York Uni­ver­si­ty School of Law. This piece is a com­men­tary on the 2018 prob­lem at the Jerome Prince Memo­r­i­al Evi­dence Moot Court Com­pe­ti­tion host­ed by Brook­lyn Law School. The issue in the prob­lem dealt with whether an inter­preter in a law enforce­ment inter­view must be avail­able for con­fronta­tion to admit an Eng­lish trans­la­tion. The views expressed in this piece do not nec­es­sar­i­ly reflect the views of the author. Rather, this arti­cle is a dis­til­la­tion of one side of the argu­ments made by the team at the Jerome Prince Memo­r­i­al Evi­dence Moot Court Competition.

2. U.S. Con­st. amend. VI

3. See Ohio v. Roberts, 448 U.S. 56, 66 (1980) (dis­cussing the need for “indi­cia of reli­a­bil­i­ty” for hearsay state­ments to be admissible).

4. See Unit­ed States v. Ushakow, 474 F.2d 1244, 1245 (9th Cir. 1973).

5. See Unit­ed States v. Nazemi­an, 948 F.2d 522 (9th Cir. 1991).

6. 541 U.S. 36, 68–69 (2011) (“Where tes­ti­mo­ni­al evi­dence is at issue, how­ev­er, the Sixth Amend­ment demands what the com­mon law required: unavail­abil­i­ty and a pri­or oppor­tu­ni­ty for cross-exam­i­na­tion. . . [This rule] applies at a min­i­mum to pri­or tes­ti­mo­ny at a pre­lim­i­nary hear­ing, before a grand jury, or at a for­mer tri­al; and to police interrogations.”).

7. 564 U.S. 647, 661 (2011) (“Accord­ing­ly, the ana­lysts who write reports that the pros­e­cu­tion intro­duces must be made avail­able for con­fronta­tion even if they pos­sess the sci­en­tif­ic acu­men of Mme. Curie and the verac­i­ty of Moth­er Tere­sa.”) (inter­nal cita­tions omitted).

8. See Unit­ed States v. Vida­cak, 553 F.3d 344, 352 (4th Cir. 2009); Unit­ed States v. Tol­liv­er, 454 F.3d 660, 664–65 (7th Cir. 2006); Unit­ed States v. Sanchez-Gondinez, 444 F.3d 957, 960 (8th Cir. 2006); Unit­ed States v. Romo-Chavez, 681 F.3d 955, 959 (9th Cir. 2012); see also Unit­ed States v. Orm Hieng, 679 F.3d 1131, 1149 (9th Cir. 2012) (Berzon J., concurring).

9. See Unit­ed States v. Charles, 722 F.3d 1319, 1327 n. 9 (11th Cir. 2013) (“[W]e find unper­sua­sive the Ninth Cir­cuit’s use of the lan­guage con­duit the­o­ry and its under­ly­ing fac­tu­al con­sid­er­a­tions to con­clude that the inter­preter and defen­dant are iden­ti­cal for tes­ti­mo­ni­al purposes.”)

10. Guan Lee v. Unit­ed States, 198 F. 596, 601 (7th Cir. 1912).

11. See Ushakow, 474 F.2d at 1245; Unit­ed States v. Da Sil­va, 725 F.2d 828, 831 (2nd Cir. 1983).

12. See Roberts, 448 U.S. at 66.

13. See Craw­ford, 541 U.S. at 60.

14. Id. at 61.

15. Id. at 68–69.

16. Id.

17. Id. at 51.

18. See Davis v. Wash­ing­ton, 547 U.S. 813, 822 (2006).

19. Unit­ed States v. Bak­er, 432 F.3d 1189, 1204 (11th Cir. 2005); see also Unit­ed States v. Arbo­laez, 450 F.3d, 1283, 1291 (11th Cir. 2006).

20. 564 U.S. 647, 661 (2011).

21. See id. at 662.

22. Id.

23. Unit­ed States v. Da Sil­va, 725 F.2d 828, 831 (2nd Cir. 1983).

24. See Unit­ed States v. Mar­tinez-Gay­tan, 213 F.3d 890, 892 (5th Cir. 2000); Orm Hieng, 679 F.3d at 1140.

25. See Unit­ed States v. Nazemi­an, 948 F.2d 522, 527 (9th Cir. 1991).

26. See Vida­cak, 553 F.3d at 352); Tol­liv­er, 454 F.3d at 664–65; Sanchez-Gondinez, 444 F.3d at 960; Romo-Chavez, 681 F.3d at 959.

27. 679 F.3d at 1140.

28. Id.

29. Id.

30. 725 F.2d at 832.

31. Id.

32. 722 F.3d at 1327.

33. Id.

34. Id. at n. 9.

35. Id. at 1328 n. 10 (“As Craw­ford instructs, a prop­er Con­fronta­tion Clause analy­sis does not begin or end with a deter­mi­na­tion of whether a state­ment con­sti­tutes ‘imper­mis­si­ble hearsay.’ Instead, a prop­er analy­sis first requires a deter­mi­na­tion of whether the declar­an­t’s state­ment is ‘tes­ti­mo­ni­al,’ i.e. a dec­la­ra­tion offered for the pur­pose of prov­ing some fact to be used at tri­al, and if so, the Sixth Amend­ment is sat­is­fied only if the declar­ant is unavail­able and there was a pri­or oppor­tu­ni­ty for cross-examination.”).

36. See Craw­ford, 541 U.S. at 51.

37. Charles, 722 F.3d at 1324.

38. Tay­lor v. State, 226 Md. App. 317, 349 (2016).

39. See Charles, 722 F. 3d at 1324 (“Charles is the declar­ant of her out-of-court Cre­ole lan­guage state­ments and the lan­guage inter­preter is the declar­ant of her out-of-court Eng­lish lan­guage statements.”).

40. Id. at 1327.

41. See Nazemi­an, 948 F.2d at 527.

42. 564 U.S. 647 (2011).

43. Id. at 658–59 (find­ing that “[a]n analyst’s cer­ti­fi­ca­tion pre­pared in con­nec­tion with a crim­i­nal inves­ti­ga­tion or pros­e­cu­tion” is tes­ti­mo­ni­al evi­dence and there­fore that “ana­lysts who write reports that the pros­e­cu­tion intro­duces must be made avail­able for con­fronta­tion” as tes­ti­mo­ni­al declarants).

44. Orm Hieng, 679 F.3d at 1149 (Berzon J., con­cur­ring) (empha­sis in original).