Separate opinion of Judge Abraham

Document Number
124-20121119-JUD-01-02-EN
Parent Document Number
124-20121119-JUD-01-00-EN
Document File
Bilingual Document File

730

SEPARATE OPINION OF JUDGE ABRAHAM

[Translation]

Disagreement with the reasoning, not with the operative part— Essential first
to consider the 1928 Treaty in order to determine whether it settles the question of
the sovereignty of the islands still in dispute — Failure to rule on the interpretation
of the 1928 Treaty — No valid justification for failing to do so — Application of
the traditional equidistance method at the very least inappropriate in t▯his
instance — Impossible to construct a provisional median line which takes accou▯nt
of all the “relevant coasts” — Inadequacy in this case of the notions of “adjustment”
or “shifting” of the provisional line.

1. I voted in favour of all the points in the operative part of the Judg -
ment. Nevertheless, I disagree with certain aspects of the Court’s reason -
ing. This opinion does not seek to criticize the reasoning as a whole, nxor
even its fundamental logic, but rather two of its individual elements. Txhey

are, first, the conclusion which the Judgment draws — or rather, in my
view, does not draw — from its consideration of the 1928 Treaty at the
end of Part 2, subsection A (paras. 40 to 56) ; and second, how the Judg -
ment deals with the issue of the construction of a “median line” axs the

first stage in the delimitation process (paras. 184 to 199).
The reasons why I disagree with those two points are as follows.

* * *

I. The Consideration of thxe 1928Treaty
as Title to Sovereigntyx over the Islands in Dispxute

2. In support of their opposing claims to sovereignty over the islands
in dispute, the Parties put forward three main series of arguments : the
first, which was essentially invoked by Colombia, was based on the bilxat -
eral 1928 Treaty and its 1930 Protocol ; the second was based on the prin-

ciple of uti possidetis juris ; and the third was based on the post-colonial
effectivités.
3. The Judgment begins by considering the issue of the 1928 Treaty.
This is fully justified : not so much by the fact that Colombia relied prin -
cipally on that Treaty as the source of its sovereignty and only advancexd

the other two series of arguments as alternatives ; but above all because
the conventional title, if its existence was established, would take prece -
dence over any other consideration, and would make the examination of
the other bases put forward by the Parties not only pointless but legallxy
impossible.

110

6 CIJ1034.indb 217 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 731

4. In other words, this was not one of those situations — which do
occur — in which the Court could consider the various legal bases pleaded

for resolving the dispute, and choose the one which it regarded as constxitut -
ing the most robust and most appropriate basis for its reasoning. It wasx
bound to examine the issue of the Treaty first and was only entitled txo move
on to consider the uti possidetis juris and the effectivités if and to the extent
that the Treaty did not accord sovereignty over the islands in dispute txo one

or other of the Parties. Indeed, if the Treaty were construed as accordixng
sovereignty to one Party, then that Party should be declared to be in poxsses -
sion of it at the present time, even if the examination of the uti possidetis and
the effectivités were to lead to conclusions in favour of the other Party. If the
1928 Treaty did derogate from the division of sovereignty over the islands
which was established by the principle of uti possidetis juris, then it was legit

imate for it to do so ; the effectivités subsequent to the Treaty could not,
whatever their nature, take precedence over the conventional title. Only a
new treaty or an agreement binding the Parties could have contradicted txhe
1928 Treaty on the question of sovereignty over the islands in dispute,
assuming that this question was settled — in whole or in part — by that lat -

ter Treaty; but no one has alleged that such a post-1928 agreement exists.
5. It was therefore crucial to determine whether the 1928 Treaty (with
its 1930 Protocol) settled the question of sovereignty over the islands cur -
rently in dispute. Moreover, it is clear that the 2007 Judgment on the
Preliminary Objections raised by Colombia did not rule on that point.

That Judgment merely noted that Article I of the 1928 Treaty expressly
accorded sovereignty to Colombia over the three islands mentioned by
name therein (San Andrés, Providencia and Santa Catalina) — which was
why the Court did not have jurisdiction over that part of the dispute,
since it had been settled by an agreement between the Parties — but that,
on the other hand, it was not easy, prima facie, to determine the other x

disputed features over which sovereignty was attributed to Colombia
under the Treaty, and that the Court did indeed have jurisdiction over
that part of the dispute, which had to be decided on the merits in the
subsequent phase of the proceedings. That was the Court’s task in thex
present Judgment.

6. Up to this point in my reasoning I have no objections to the Judg -
ment.
In paragraph 42, after noting that, under the terms of the 1928 Treaty,
Colombia has sovereignty over “San Andrés, Providencia and Santa
Catalina and over the other islands, islets and reefs forming part of thxe

San Andrés Archipelago”, the Court is right to deduce that : “in order to
address the question of sovereignty over the maritime features in disputxe,
[it] needs first to ascertain what constitutes the San Andrés Archixpelago”.
In the context of this paragraph, the word “first” means that thxe question
thus formulated needed to be resolved before the Court turned — but
only if that were still to be necessary after answering the first quesxtion —

to the consideration of the other arguments of the Parties, based on the
uti possidetis juris and the effectivités.

111

6 CIJ1034.indb 219 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 732

7. It is clear, however, that at the end of the examination which it con -
ducts in paragraphs 52 to 55, the Court does not do what it said it would

do in paragraph 42 : it does not “ascertain what constitutes the San
Andrés Archipelago”. In fact it does not draw any conclusion and mxerely
notes that, since it cannot reach a definitive decision on the scope oxf the
1928 Treaty concerning the features in dispute, it can only settle the dis -
pute over sovereignty on the basis of the arguments of the Parties “wxhich

are not based on the composition of the Archipelago under the 1928
Treaty” — that is to say, the arguments concerning the uti possidetis juris
and the effectivités (Judgment, para. 56). It then moves on to consider
those other arguments.
8. In so doing, in my opinion, the Court commits a serious legal error :
it fails, without valid justification, to rule on the interpretation oxf the

1928 Treaty, and, more specifically, on the meaning of the words “over
the other islands, islets and reefs forming part of the San Andrés Arxchi -
pelago”, within the meaning of Article I of the Treaty.
9. The fundamental reasons for this failure are provided in para -
graph 53 of the Judgment :

“the question about the composition of the Archipelago cannot. . . be
definitively answered solely on the basis of the geographical locationx

of the maritime features in dispute or on the historical records relatinxg
to the composition of the San Andrés Archipelago referred to by the
Parties, since this material does not sufficiently clarify the matter”x.

In essence, the Court notes that when the 1928 Treaty refers to the
“San Andrés Archipelago” it does not define its composition ; that the
sole fact that some islands lie close to the main island of San Andréxs is
insufficient to conclude that they form part of the Archipelago, whereaxs
other, more distant, islands do not — since it would be necessary to deter -

mine a cut-off point for establishing appurtenance to the Archipelago,
which the Treaty does not allow ; and, finally, that the examination of the
documents communicated to the Court by the Parties, which were meant
to shed light on the context in which the Treaty was negotiated and con -
cluded, does not establish with any certainty what the Parties intended x

the reference to the “San Andrés Archipelago” to signify at thex time.
10. None of the aforementioned reasons justifies the Court’s failure tox
interpret the Treaty : they merely emphasize that the Treaty is unclear on
this point, identify the difficulties encountered when seeking to define its
meaning and scope, and indicate that it is impossible to draw a definixtive

conclusion. None of that justifies the Court’s failure to interpretx the
Treaty, whose meaning is disputed by the Parties. All that can be deducexd
from the reasons given by the Court is that the interpretation is diffixcult
in this case. True. But the difficulty of interpreting a legal text is xnot — is
never — a valid reason for a failure to do so by the court which is respon -
sible for applying it. A text’s obscurity is a sign that it needs to xbe inter -

preted, never an obstacle to that interpretation. The court may not be
certain about the meaning of the text, it may hesitate over the solutionx to

112

6 CIJ1034.indb 221 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 733

adopt; that is not unusual. But it is the court’s duty to decide, irrespecxtive
of its doubts — doubts which it is moreover perfectly entitled to express

at the very moment when it does decide.
11. Admittedly, there are cases when, faced with a relatively obscurely
worded norm, the court prefers to avoid coming down in favour of one
particular questionable interpretation, and decides to set the difficulxty
aside and settle the dispute on the basis of other legally relevant and xsuf -

ficient considerations. That is the mark of healthy judicial caution. xHow-
ever, it still has to be legally possible, given the particular facts ofx the
case, to rule without establishing the meaning of the norm whose scope ixs
in doubt. That is not always the case. For example, it is not the case ixn
this instance, for the reasons which I have set out above : the 1928 Treaty,
the uti possidetis and the effectivités are not alternative legal bases, which

are on an equal footing, and between which the Court could choose in
order to settle the issue of sovereignty. It was necessary first to dextermine
the effects of the 1928 Treaty on sovereignty before the rest could — if
appropriate — be examined. Deciding cannot mean merely noting that
the task is difficult : the Court has not done its duty.

12. Admittedly, when it writes, in paragraph 56, that, in order to
resolve the dispute, it must examine the arguments of the Parties which x
are not based on the Treaty, whose meaning it regards as being in doubt,x
the Court already knows that when it considers the effectivités it will find

sufficient robust and relatively uncontentious evidence on which to basxe a
conclusion in favour of awarding sovereignty to one of the Parties.
However, that does not alter the problem. For the reasons which I have
already stated, the Court was not at leisure to choose between the Treatxy
and the effectivités on the basis of which of the two grounds appeared to
be the more robust.

Moreover, if the Parties had pleaded solely on the basis of the Treaty, x
the Court would certainly not have evaded its duty of interpretation, thxe
performance of which may be difficult but is never impossible.
13. I would add, to anticipate a possible objection, that a court’s duty
to interpret a treaty which has been adduced by a party, when it is not

legally possible to rely on a strictly alternative basis, is not limitedx to cases
in which the provision invoked seeks to define a rule of a general andx
impersonal nature, a genuine norm, that is to say, one which is abstract
and permanent. The duty to interpret is equally applicable in those casexs,
like the present one, in which the contentious clause confers a specific title

on a party, notably a title to sovereignty. In such cases there is no rexason
to derogate from the fundamental principle that a court is not entitled to
cite the obscurity of the treaty as justification for not interpretingx it. I
regret that the Court disregarded that principle in this case.
14. Having said that, I think that the Court’s final conclusion would
have been the same if it had proceeded as it ought to have done.

15. It would first have noted that, unless the last clause of the first xpara
graph of Article I of the 1928 Treaty were rendered ineffective, it must

113

6 CIJ1034.indb 223 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 734

inevitably be acknowledged that at least some of the features in disputex in
the present phase of the proceedings belong to Colombia on the basis of

the Treaty, since they form part of the “San Andrés Archipelago”x. That
provision in fact implies that islands other than San Andrés, Providencia
and Santa Catalina form part of the “San Andrés Archipelago” under the
Treaty, and those other islands can be none other than those which are
presently in dispute, or certain of them at least. Nicaragua’s positixon, that

“the Archipelago comprises only the islands of San Andrés, Providexncia
and Santa Catalina” (para. 48 of the Judgment) is incompatible with the
Treaty, since it renders it meaningless. A simple glance at the map is sxuf -
ficient to conclude — once you disregard all the islands to the west of the
82º W meridian, which the 1930 Protocol declares not to belong to the
Archipelago under the Treaty — that the Archipelago includes at least the

Alburquerque Cays and the East-Southeast Cays, which lie closest to
San Andrés. Those islands therefore definitely belong to Colombia underx
the Treaty, and the Court ought to have noted that fact, instead of cau -
tiously indicating that “given their geographical location” they “xcould be
seen as forming part of the Archipelago” (ibid., para. 53), before adding

that this geographical criterion was not decisive.
16. In my opinion, there is sufficient evidence to consider that in 1928
the islands of Roncador, Quitasueño and Serrana were also regarded as
forming part of the San Andrés Archipelago, but it is not necessary to
settle that question, since the second paragraph of Article I of the Treaty

expressly precludes sovereignty over those three features from being
attributed to Colombia. The fact that the reason given is that their appxur -
tenance was in dispute between Colombia and the United States of Amer-
ica at the time, a dispute which subsequently disappeared when the Unitexd
States renounced its claim, does not alter the indisputable fact that thxe
1928 Treaty does not in itself confer a title of sovereignty on Colombia

over the three features in question. The Court was therefore able to leaxve
the issue unresolved of whether Roncador, Quitasueño and Serrana
formed part of the San Andrés Archipelago in the sense in which the two
States understood that notion in 1928.
17. Finally, it seems to me that Bajo Nuevo and Serranilla are too far

away from San Andrés to be reasonably regarded, at first sight, as form -
ing part of the Archipelago, and that this assumption must be made,
unless there is sufficiently convincing evidence to the contrary in thex
travaux préparatoires of the 1928 Treaty. However, Colombia did not
provide any such evidence in support of its claim.

18. I therefore conclude that the 1928 Treaty accords Colombia sover -
eignty not only over San Andrés, Providencia and Santa Catalina (sinxce
these three islands are no longer at issue in the present phase of the pro -
ceedings), but also over the Alburquerque Cays and the East-Southeast
Cays; however, it does not accord either of the two Parties sovereignty
over the other maritime features in dispute.

19. In respect of the latter — Quitasueño, Serrana, Roncador, Serra -
nilla and Bajo Nuevo — but of them alone, the Court had to move to the

114

6 CIJ1034.indb 225 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 735

examination of the arguments based on the uti possidetis and the
post-colonial effectivités. In this regard, I support the Judgment’s subse -

quent reasoning: a title in favour of one or other of the Parties cannot be
established on the basis of the principle of uti possidetis ; the effectivités
are in Colombia’s favour.
20. Ultimately, my serious reservations about the reasoning in the
Judgment did not prevent me from voting in favour of point 1 of the

operative part, since my conclusion is the same as that of my colleaguesx.

* * *

II. The Construction of a Prxovisional Median Line
as the First Stage in thex Method for Fixing
the Maritime Boundary

21. As far as the maritime delimitation is concerned, my disagreement

relates less to what the Court has done — moreover, I agree with the end
result of the process, and I voted in favour of points 4 and 5 of the oper -
ative part — than to how it is presented, which appears to me to be
largely fallacious. In short, my opinion is that, although the Court staxtes
that it is following the traditional method, as described in particular xin its

Judgment in the case between Romania and Ukraine (Maritime Delimita ‑
tion in the Black Sea (Romania v. Ukraine), Judgment, I.C.J. Reports 2009,
p. 61), in reality it diverges very considerably from it and actually it can -
not do otherwise, since it is clear that the said method is inappropriatxe in
the present case.
22. The method in question is recalled in paragraphs 190 to 193 of the

Judgment. It consists of first constructing a provisional median line,x that
is to say, a line which is at equal distance from the opposite coasts ofx the
two States which generate entitlements to overlapping maritime spaces —
those overlapping entitlements being the very reason why it is necessaryx
to effect a delimitation. Where the relevant coasts are adjacent, the provi -

sional line is termed an equidistance line, but that does not make any
substantial difference and moreover is not the case here. The second sxtage
is to adjust or shift the provisional line thus obtained in order to take
account of any particular circumstances which might require the line to x
be adjusted or shifted in order to achieve an equitable solution. Finallxy,

in a third stage, the Court must check that the maritime areas awarded txo
the Parties by virtue of the delimitation obtained at the end of the previ -
ous stage are not markedly disproportionate to their respective relevantx
coasts — the coasts which generate the entitlements to the overlapping
spaces.

23. The Court considers the arguments by which Nicaragua sought to
convince it that the said method was inappropriate in the present case oxn

115

6 CIJ1034.indb 227 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 736

the grounds that the particular geographical situation was one in which x
the Court should not begin by constructing a provisional median line. Itx

acknowledges that the “three-stage process is not . . . to be applied in a
mechanical fashion” and “that it will not be appropriate in every xcase to
begin with a provisional equidistance/median line” (Judgment, para. 194).
However, it dismisses Nicaragua’s arguments and states that, althoughx
there are undoubtedly particular circumstances which justify adjusting

the provisional median line, there is no reason not to begin by construcxt -
ing such a line nor to use it as a starting-point for the delimitation. xCon-
sequently, the Court affirms that it will adhere to its “standard mexthod”
(ibid., para. 199), and it proceeds to do so — or rather it claims to pro -
ceed to do so — in paragraphs 200 to 204 (first stage : construction of the
provisional median line), in paragraphs 205 to 238 (second stage : adjust -

ment or shifting of the provisional line), and in paragraphs 239 to 247
(third stage: disproportionality test).

24. Nevertheless, it is obvious that the construction of a provisional
median line as a starting-point for the delimitation is not only highly

inappropriate in this case, but that it is even virtually impossible.

25. The reason for this is very simple. The overlapping entitlements
which make the delimitation necessary in this instance do not exist
because two opposite (or adjacent) coasts are generating projections

which overlap in an intermediate area, as is usually the case. Here, thex
overlapping entitlements occur because, within the exclusive economic
zone measured from the Nicaraguan coast, there are islands belonging to x
Colombia which generate an entitlement to an exclusive economic zone
for that State in all directions. In other words, the overlapping does nxot
only occur between the Nicaraguan coast and the Colombian islands

(that is to say, in the area to the west of the Colombian islands and to the
east of the Nicaraguan coast); it also occurs in the areas to the north, east
and south of the Colombian islands — and even between them. This is
shown very clearly on sketch-map No. 7 in the Judgment (p. 687), which
depicts the “relevant maritime area”, that is to say, the area of xoverlap -

ping entitlements within which the Court is called upon to effect the x
delimitation.
26. Plainly, therefore, no “median line” can take account of the geo -
graphical reality which was submitted for the Court’s consideration, xnot
because of any “relevant particular circumstance” which would justxify the

adjustment of a provisional line without making it impossible to con -
struct it in the first place, but because of the essential facts of thxe dispute
brought before the Court, which make the very notion of a “median linxe”
meaningless in the present case.
27. It is admittedly possible to construct a line which is equidistant
from the Nicaraguan coast and the west-facing coasts of the Colombian

islands, and that is what the Court does, affirming that in so doing itx has
completed the first stage of its “standard method”. But a glancex at that

116

6 CIJ1034.indb 229 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 737

line, which is shown on sketch-map No. 8 (p. 701), is sufficient to realize
that it is “median” in name only : it may be equidistant from the Nicara -

guan coast (more precisely from the Nicaraguan islands adjacent to that
State’s mainland coast), on its western side, and the western coasts of the
Colombian islands, on its eastern side. However, it does not take any
account — indeed the manner of its construction means that it cannot
take any account — of the entire area to the east of the Colombian

islands, which nonetheless also forms part of the overlapping area. Thisx is
not a “particular circumstance” which would justify a subsequent axdjust -
ment or shifting of the line. It is a fundamental defect which deprives xthe
line of its alleged “median” character. This can be explained by ax specific
characteristic of the case : the Court could only construct that line by
taking base points, as far as Colombia is concerned, which were located x

exclusively on the west-facing coasts of the islands belonging to
that State. It could not adopt any base points on the east-, north- and
south-facing coasts of those islands since they do not face the Nicaraguan
coast. However, as I recalled above, all of the coasts of Colombia’s x
islands, not just the west-facing parts of those coasts, generate

entitlements to an exclusive economic zone which overlap with those of
Nicaragua.
28. In other words, in order to be able to construct a line which has at
least the semblance of a “median line” — although in my view even that
is debatable — the Court deliberately had to ignore the majority of

Colombia’s relevant coasts. However, in order to perform its designated
function in the delimitation process, a median line must take into accouxnt
all the “relevant coasts” of the States present, that is to say, axll the coasts
which generate the projections creating the overlapping entitlements
which make the delimitation necessary.
29. The Judgment itself recalls that point in paragraph 191: the median

line has to be “constructed using the most appropriate base points onx the
coasts of the Parties”. These points are admittedly chosen, but they xcan -
not be chosen in just any way : in order for them to be “the most appro -
priate”, they must take satisfactory account of all the “relevant xcoasts”
and not just one part of those coasts. However, as far as the Colombian x

islands are concerned, the Judgment rightly points out that the relevantx
coasts constitute “the entire coastline of these islands, not merely xthe
west-facing coasts” (para. 151). This suggests that a median line corre -
sponding to the definition in the “standard method” would have to be
constructed from base points on all the coasts of the Colombian islands,

and not only on their west-facing parts. Clearly, however, that is not pxos -
sible in this case.

30. Instead of concluding from this that the construction of a median
line — albeit a provisional one — is at the very least inappropriate, if not
impossible, in this case, the Court decides to construct one all the samxe

without taking into account (simply because it cannot) the majority ofx the
coasts of the Colombian islands. In so doing, it appears to forget in paxra -

117

6 CIJ1034.indb 231 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 738

graphs 200 to 204, in which it selects base points which, on the Colom -
bian side, are located exclusively on the western sides of the islands, xwhat

it explained in paragraphs 151 and 191.
31. It is true that this enables it to construct a line (depicted on
sketch-map No. 8). But that line is only “median” with respect to one
part of the “relevant area” to delimit (the area shown on sketch-map
No. 7); it is otherwise entirely meaningless. In my view, therefore, the line x

constructed cannot be regarded as a “median line”, that is, as an xaccept -
able starting-point for the delimitation, which will subsequently only be
adjusted or shifted to a necessarily limited extent, in order to take acxcount
of particular circumstances.

32. Moreover, further on in its reasoning the Judgment implicitly

acknowledges that fact, in two ways.
First, after adjusting the provisional line by shifting it considerably x
eastwards (in order, therefore, to move it closer to the Colombian islaxnds),
the Court notes that even after that adjustment the result would not be x
equitable if the line “extend[ed]... into the parts of the relevant axrea north

of point 1 or south of point 5”, that is to say, to the north and south of
the principal Colombian islands, and that furthermore the line in ques -
tion would cut off Nicaragua from the areas to the east of those islanxds,
areas “into which the Nicaraguan coast projects” (Judgment, para.x 236).
That is perfectly true, but does it not constitute an acknowledgment thaxt

the provisional line is not fit for purpose, with regard to a large paxrt of
the area in which the delimitation is to be effected, that is to say, xall the
sectors to the north, south and east of the principal Colombian islands ?

Second, and as a consequence of the foregoing, the Court is induced to
construct two horizontal lines along lines of latitude passing to the noxrth

through point 1 (which is located to the north of Santa Catalina, and
approximately level with Roncador) and to the south through point 9
(which is located level with East-Southeast Cays) with a view to delimit -
ing the area to the east of the Colombian islands (ibid., para. 237). How -
ever, it is difficult to regard these two horizontal lines as a mere

“adjustment” or even “shifting” of the provisional line. With the excep -
tion of the starting-point of the first line, those lines are actually entirely
unrelated to the provisional line. The same goes for the addition of no x
fewer than four maritime frontier points (points 6 to 9 on sketch-map
No. 11, p. 714) in the southern part of the area to be delimited which,

rather than adjusting or shifting the provisional line, are in fact suppxle -
mentary to it.
33. In short, after describing as a “median line” a line which does noxt
really merit that description, the Court terms an “adjustment” or x“shift”
a process which does not really merit being termed as such. Perhaps the x
Judgment envisaged that process (the construction of two horizontal linxes

and the fixing of points 6 to 9) as a separate stage after the adjustment or
shifting of the line. But if so, would that not be adding another stage —

118

6 CIJ1034.indb 233 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 739

and a decisively important one in this case — to the “traditional method”

(or “standard method”) to which the Court nevertheless promised xto
adhere in paragraph 199 ?
34. I do not wish to say that the Court was wrong to delimit the spaces
constituting the relevant area in the way that it did. On the contrary, xI
think that it adopted the most reasonable solution, and that each stage xin

its construction was intrinsically justified. However, my opinion is txhat it
would have been clearer and more honest of the Court to acknowledge
that it could not follow the so-called “standard” method in this cxase
because the geographical framework did not at all lend itself to the appli -
cation of that method. In this instance it thus found itself in the situxation

in which “compelling reasons . . . in the particular case” made it unfeasi -
ble to construct the provisional median line (Maritime Delimitation in the
Black Sea (Romania v. Ukraine), Judgment, I.C.J. Reports 2009, p. 101,
para. 116) or, at the very least, in one in which the application of the
equidistance method was “inappropriate” (Territorial and Maritime Dis ‑

pute between Nicaragua and Honduras in the Caribbean Sea (Nicaragua v.
Honduras), Judgment, I.C.J. Reports 2007 (II), p. 741, para. 272, men -
tioned in paragraph 194 of the present Judgment).
35. I understand that the Court wishes to give all its observers, and
first and foremost States, the impression that it does not use arbitraxry

methods to achieve an equitable solution, but that it implements proven x
and consistent techniques. And it is perfectly true that there is nothinxg
arbitrary about the Court’s approach, which is characterized merely bxy a
scrupulous search for the best solution. However, there are cases which x
are presented in such specific terms that it is, on the whole, preferaxble to

acknowledge that the Court needs to depart from its usual technique, and
to explain why, rather than to sacrifice clarity and intelligibility txo the
semblance of an illusory continuity.

(Signed) Ronny Abraham.

119

6 CIJ1034.indb 235 7/01/14 12:43

Bilingual Content

730

OPINION INDIVIDUELLE DE M. LE JUGE ABRAHAM

Désaccord avec le raisonnement, pas avec le dispositif — Nécessité d’examiner
en premier lieu le traité de 1928 pour déterminer s’il permet d▯e régler la question
de la souveraineté des îles encore en litige — Défaut de statuer sur l’interprétation
du traité de 1928 — Absence de justification valable pour s’abstenir de le faire —
Application de la méthode traditionnelle de l’équidistance à tout le moins
inappropriée en l’espèce — Impossibilité de construire une ligne médiane provisoire
tenant compte de l’ensemble des « côtes pertinentes » — Caractère inadéquat en
l’espèce des notions d’« ajustement» ou de « déplacement» de la ligne provisoire.

1. J’ai voté en faveur de tous les points du dispositif de l’arrêxt. J’ai
cependant des désaccords sur certains aspects du raisonnement suivi pxar
la Cour. La présente opinion ne vise pas à critiquer la motivationx dans
son ensemble, ni même ses articulations essentielles, mais deux de ses élé -

ments particuliers. Il s’agit, d’une part, de la conclusion que l’xarrêt tire
— ou plutôt, à mon sens, ne tire pas — de son examen du traité de 1928,
au terme de la sous-section 2.A de la section II (par. 40-56) ; d’autre part,
du traitement par l’arrêt de la question du tracé d’une «x ligne médiane »

en tant que première étape du processus de délimitation (par. 184-199).
Les raisons de mon désaccord sur ces deux points sont les suivantes.

* * *

I. L’examen du traité de 1928 x
en tant que titre de soxuveraineté sur les îlexs en litige

2. Les Parties ont invoqué, au soutien de leurs prétentions contrairexs à
la souveraineté sur les îles en litige, trois séries principalexs d’argument: la
première, invoquée essentiellement par la Colombie, tirée du trxaité bilaté -
ral de 1928 et de son protocole de 1930 ; la deuxième tirée du principe de

l’uti possidetis juris ; la troisième tirée des effectivités postcoloniales.

3. L’arrêt commence par examiner la question du traité de 1928. Cexla
est pleinement justifié : non pas tellement par le fait que la Colombie a
invoqué le traité, comme fondement de sa souveraineté, à titxre principal,

et n’a mis en avant les deux autres séries d’arguments qu’à titre subsi -
diaire; mais surtout parce que le titre conventionnel, si son existence était
établie, l’emporterait sur toute autre considération, et rendraxit l’examen
des autres bases invoquées par les Parties non seulement inutile maisx juri -
diquement impossible.

110

6 CIJ1034.indb 216 7/01/14 12:43 730

SEPARATE OPINION OF JUDGE ABRAHAM

[Translation]

Disagreement with the reasoning, not with the operative part— Essential first
to consider the 1928 Treaty in order to determine whether it settles the question of
the sovereignty of the islands still in dispute — Failure to rule on the interpretation
of the 1928 Treaty — No valid justification for failing to do so — Application of
the traditional equidistance method at the very least inappropriate in t▯his
instance — Impossible to construct a provisional median line which takes accou▯nt
of all the “relevant coasts” — Inadequacy in this case of the notions of “adjustment”
or “shifting” of the provisional line.

1. I voted in favour of all the points in the operative part of the Judg -
ment. Nevertheless, I disagree with certain aspects of the Court’s reason -
ing. This opinion does not seek to criticize the reasoning as a whole, nxor
even its fundamental logic, but rather two of its individual elements. Txhey

are, first, the conclusion which the Judgment draws — or rather, in my
view, does not draw — from its consideration of the 1928 Treaty at the
end of Part 2, subsection A (paras. 40 to 56) ; and second, how the Judg -
ment deals with the issue of the construction of a “median line” axs the

first stage in the delimitation process (paras. 184 to 199).
The reasons why I disagree with those two points are as follows.

* * *

I. The Consideration of thxe 1928Treaty
as Title to Sovereigntyx over the Islands in Dispxute

2. In support of their opposing claims to sovereignty over the islands
in dispute, the Parties put forward three main series of arguments : the
first, which was essentially invoked by Colombia, was based on the bilxat -
eral 1928 Treaty and its 1930 Protocol ; the second was based on the prin-

ciple of uti possidetis juris ; and the third was based on the post-colonial
effectivités.
3. The Judgment begins by considering the issue of the 1928 Treaty.
This is fully justified : not so much by the fact that Colombia relied prin -
cipally on that Treaty as the source of its sovereignty and only advancexd

the other two series of arguments as alternatives ; but above all because
the conventional title, if its existence was established, would take prece -
dence over any other consideration, and would make the examination of
the other bases put forward by the Parties not only pointless but legallxy
impossible.

110

6 CIJ1034.indb 217 7/01/14 12:43 731 différend territoriaxl et maritime (op. ind. axbraham)

4. En d’autres termes, nous ne sommes pas dans un cas, comme il s’en x

produit, où la Cour pouvait choisir, entre plusieurs bases juridiquesx invo-
quées devant elle pour résoudre le différend, celle qui lui pxaraissait la plus
solide et la plus opportune pour fonder son raisonnement. Elle était xtenue
d’examiner d’abord la question du traité, et n’avait le droixt de passer à
l’examen de l’uti possidetis juris et des effectivités que si et dans la mesure

où le traité n’attribuait pas la souveraineté sur les îlexs en litige à l’une ou
l’autre des Parties. En effet, si le traité devait être interxprété comme attri -
buant la souveraineté à une Partie, celle-ci devrait être déclarée la possé -
der à l’heure actuelle, quand bien même l’examen de l’uti possidetis et des
effectivités conduirait à des conclusions favorables à l’axutre Partie. Le

traité de 1928 a valablement pu déroger, s’il l’a fait, au pxartage de la sou -
veraineté sur les îles résultant du principe de l’uti possidetis juris ; les effec-
tivités postérieures au traité ne sauraient, quelles qu’ellexs soient, l’emporter
sur le titre conventionnel. Seul un nouveau traité ou un accord liantx les
Parties aurait pu contredire le traité de 1928 sur la question de la sou-

veraineté sur les îles en litige, à supposer que cette questionx ait été réglée
— en tout ou en partie — par ce dernier traité ; mais l’existence d’un tel
accord postérieur à 1928 n’a été alléguée par persoxnne.
5. Il était donc crucial de déterminer si le traité de 1928 (avecx son pro -
tocole de 1930) réglait la question de la souveraineté sur les îles présente -

ment en litige. Par ailleurs, il est clair que l’arrêt rendu en 2007 sur les
exceptions préliminaires soulevées par la Colombie ne s’est pasx prononcé
sur ce point. Cet arrêt s’est borné à constater que l’artxicle 1 erdu traité de
1928 attribuait expressément souveraineté à la Colombie sur lesx trois îles
qui y sont nommément désignées (San Andrés, Providencia, Santa Cata -

lina) — raison pour laquelle la Cour n’était pas compétente pour cxonnaître
de cette partie du différend, celle-ci ayant été réglée par un accord entre
les Parties — mais qu’en revanche il n’était pas évident, prima facie, de
déterminer sur quelles autres des formations en litige le traité attribuait
souveraineté à la Colombie, et que cette partie du différend xrelevait bien

de la compétence de la Cour et devait être tranchée au fond dans la phase
suivante de la procédure. C’est cette tâche qu’il appartenaixt à la Cour
d’accomplir dans le présent arrêt.
6. Jusqu’à ce point de mon raisonnement, je n’ai rien à objectexr à l’ar -
rêt.

C’est à bon droit que la Cour, au paragraphe 42, après avoir noté que,
aux termes du traité de 1928, la Colombie a la souveraineté sur «x les îles de
San Andrés, de Providencia, de Santa Catalina, et sur les autres îles, îlots et
récifs qui font partie de l’archipel de SanAndrés», en déduit que, «pour se
prononcer sur la question de la souveraineté sur les formations maritximes

en litige, [elle] doit … d’abord établir quelles sont les formations qui consti -
tuent l’archipel de SanAndrés». Le mot «d’abord» signifie, dans le contexte
de ce paragraphe, que la question ainsi formulée doit être résoxlue avant que
la Cour en vienne — mais seulement si cela était encore nécessaire après
qu’il aura été répondu à la première question— à l’examen des autres argu -

ments des Parties, fondés sur l’uti possidetis juris et les effectivités.

111

6 CIJ1034.indb 218 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 731

4. In other words, this was not one of those situations — which do
occur — in which the Court could consider the various legal bases pleaded

for resolving the dispute, and choose the one which it regarded as constxitut -
ing the most robust and most appropriate basis for its reasoning. It wasx
bound to examine the issue of the Treaty first and was only entitled txo move
on to consider the uti possidetis juris and the effectivités if and to the extent
that the Treaty did not accord sovereignty over the islands in dispute txo one

or other of the Parties. Indeed, if the Treaty were construed as accordixng
sovereignty to one Party, then that Party should be declared to be in poxsses -
sion of it at the present time, even if the examination of the uti possidetis and
the effectivités were to lead to conclusions in favour of the other Party. If the
1928 Treaty did derogate from the division of sovereignty over the islands
which was established by the principle of uti possidetis juris, then it was legit

imate for it to do so ; the effectivités subsequent to the Treaty could not,
whatever their nature, take precedence over the conventional title. Only a
new treaty or an agreement binding the Parties could have contradicted txhe
1928 Treaty on the question of sovereignty over the islands in dispute,
assuming that this question was settled — in whole or in part — by that lat -

ter Treaty; but no one has alleged that such a post-1928 agreement exists.
5. It was therefore crucial to determine whether the 1928 Treaty (with
its 1930 Protocol) settled the question of sovereignty over the islands cur -
rently in dispute. Moreover, it is clear that the 2007 Judgment on the
Preliminary Objections raised by Colombia did not rule on that point.

That Judgment merely noted that Article I of the 1928 Treaty expressly
accorded sovereignty to Colombia over the three islands mentioned by
name therein (San Andrés, Providencia and Santa Catalina) — which was
why the Court did not have jurisdiction over that part of the dispute,
since it had been settled by an agreement between the Parties — but that,
on the other hand, it was not easy, prima facie, to determine the other x

disputed features over which sovereignty was attributed to Colombia
under the Treaty, and that the Court did indeed have jurisdiction over
that part of the dispute, which had to be decided on the merits in the
subsequent phase of the proceedings. That was the Court’s task in thex
present Judgment.

6. Up to this point in my reasoning I have no objections to the Judg -
ment.
In paragraph 42, after noting that, under the terms of the 1928 Treaty,
Colombia has sovereignty over “San Andrés, Providencia and Santa
Catalina and over the other islands, islets and reefs forming part of thxe

San Andrés Archipelago”, the Court is right to deduce that : “in order to
address the question of sovereignty over the maritime features in disputxe,
[it] needs first to ascertain what constitutes the San Andrés Archixpelago”.
In the context of this paragraph, the word “first” means that thxe question
thus formulated needed to be resolved before the Court turned — but
only if that were still to be necessary after answering the first quesxtion —

to the consideration of the other arguments of the Parties, based on the
uti possidetis juris and the effectivités.

111

6 CIJ1034.indb 219 7/01/14 12:43 732 différend territoriaxl et maritime (op. ind. axbraham)

7. Force est cependant de constater que, au terme de l’examen auquel

elle se livre dans les paragraphes 52 à 55, la Cour ne fait pas ce qu’elle a
annoncé qu’elle ferait au paragraphe 42 : elle s’abstient d’« établir quelles
sont les formations qui constituent l’archipel de San Andrés ». Elle ne tire
en effet aucune conclusion, et se borne à constater que, faute de pxouvoir
prendre parti avec certitude sur la portée du traité de 1928 relatxivement

aux formations en litige, elle ne peut régler le différend sur lxa souveraineté
que sur la base des arguments des Parties « autres que ceux fondés sur la
composition de l’archipel aux termes du traité de 1928 » — c’est-à-dire les
arguments relatifs à l’uti possidetis juris et aux effectivités (arrêt, par. 56).
Elle passe ensuite à l’examen de ces autres arguments.

8. Ce faisant, à mes yeux, la Cour commet une erreur juridique grave :
elle s’abstient, sans justification valable, de statuer sur l’inxterprétation du
traité de 1928, et plus précisément sur le sens des mots « sur les autres îles,
îlots et récifs qui font partie de l’archipel de San Andrés », tels qu’ils sont
employés à l’article 1 erdu traité.

9. Les raisons de cette abstention sont fournies, pour l’essentiel, au
paragraphe 53 de l’arrêt :

«la question de la composition de l’archipel ne peut être tranchéxe en
se fondant uniquement sur la situation géographique des formations
maritimes en litige ou sur les documents historiques invoqués à cext
égard par les Parties, faute d’indications suffisamment claires ».

En résumé, la Cour constate que le traité de 1928, tout en se référant à
l’«archipel de San Andrés », n’en définit pas la consistance ; que le seul

fait que certaines îles soient proches de l’île principale de Sxan Andrés ne
suffit pas pour conclure qu’elles font partie de l’archipel alorsx que d’autres,
plus éloignées, n’en feraient pas partie — car il faudrait déterminer le
degré de proximité à partir duquel l’appartenance à l’xarchipel serait éta -
blie, ce que le traité ne permet pas de faire ; enfin que l’examen des docu -

ments communiqués à la Cour par les Parties, et qui sont censésx éclairer
le contexte de la négociation et de la conclusion du traité, ne pexrmet pas
de parvenir à une quelconque certitude sur ce que les Parties ont vouxlu
signifier, à l’époque, par la référence à l’« archipel de San Andrés ».
10. Aucune de ces raisons ne justifie l’abstention de la Cour d’intexrpré -

ter le traité ; les motifs précités ne font que souligner l’absence de clartéx,
sur le point considéré, du traité, les difficultés auxquellxes on se heurte
lorsque l’on cherche à en définir le sens et la portée, l’xabsence d’une
conclusion certaine à laquelle on puisse parvenir. Rien de tout cela xne
justifie que le juge s’abstienne d’interpréter le traité dxont le sens est

controversé entre les Parties. Tout ce que l’on peut déduire dexs motifs
donnés par la Cour, c’est que l’interprétation est en l’exspèce difficile.
Certes. Mais la difficulté d’interpréter un texte juridique n’est pas — n’est
jamais — un motif valable de s’en abstenir pour le juge chargé de l’xappli -
quer. L’obscurité d’un texte appelle l’interprétation ; elle n’y fait jamais

obstacle. Que le juge ne soit pas certain du sens du texte, qu’il héxsite sur

112

6 CIJ1034.indb 220 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 732

7. It is clear, however, that at the end of the examination which it con -
ducts in paragraphs 52 to 55, the Court does not do what it said it would

do in paragraph 42 : it does not “ascertain what constitutes the San
Andrés Archipelago”. In fact it does not draw any conclusion and mxerely
notes that, since it cannot reach a definitive decision on the scope oxf the
1928 Treaty concerning the features in dispute, it can only settle the dis -
pute over sovereignty on the basis of the arguments of the Parties “wxhich

are not based on the composition of the Archipelago under the 1928
Treaty” — that is to say, the arguments concerning the uti possidetis juris
and the effectivités (Judgment, para. 56). It then moves on to consider
those other arguments.
8. In so doing, in my opinion, the Court commits a serious legal error :
it fails, without valid justification, to rule on the interpretation oxf the

1928 Treaty, and, more specifically, on the meaning of the words “over
the other islands, islets and reefs forming part of the San Andrés Arxchi -
pelago”, within the meaning of Article I of the Treaty.
9. The fundamental reasons for this failure are provided in para -
graph 53 of the Judgment :

“the question about the composition of the Archipelago cannot. . . be
definitively answered solely on the basis of the geographical locationx

of the maritime features in dispute or on the historical records relatinxg
to the composition of the San Andrés Archipelago referred to by the
Parties, since this material does not sufficiently clarify the matter”x.

In essence, the Court notes that when the 1928 Treaty refers to the
“San Andrés Archipelago” it does not define its composition ; that the
sole fact that some islands lie close to the main island of San Andréxs is
insufficient to conclude that they form part of the Archipelago, whereaxs
other, more distant, islands do not — since it would be necessary to deter -

mine a cut-off point for establishing appurtenance to the Archipelago,
which the Treaty does not allow ; and, finally, that the examination of the
documents communicated to the Court by the Parties, which were meant
to shed light on the context in which the Treaty was negotiated and con -
cluded, does not establish with any certainty what the Parties intended x

the reference to the “San Andrés Archipelago” to signify at thex time.
10. None of the aforementioned reasons justifies the Court’s failure tox
interpret the Treaty : they merely emphasize that the Treaty is unclear on
this point, identify the difficulties encountered when seeking to define its
meaning and scope, and indicate that it is impossible to draw a definixtive

conclusion. None of that justifies the Court’s failure to interpretx the
Treaty, whose meaning is disputed by the Parties. All that can be deducexd
from the reasons given by the Court is that the interpretation is diffixcult
in this case. True. But the difficulty of interpreting a legal text is xnot — is
never — a valid reason for a failure to do so by the court which is respon -
sible for applying it. A text’s obscurity is a sign that it needs to xbe inter -

preted, never an obstacle to that interpretation. The court may not be
certain about the meaning of the text, it may hesitate over the solutionx to

112

6 CIJ1034.indb 221 7/01/14 12:43 733 différend territoriaxl et maritime (op. ind. axbraham)

la solution à adopter, ce sont là des circonstances qui ne sont paxs rares.

Mais l’office du juge est de trancher, quels que soient ses doutes — qu’il
ne lui est d’ailleurs pas interdit d’exprimer au moment même oùx il décide.
11. Il arrive certes que, en présence d’une norme formulée de façxon
plus ou moins obscure, le juge préfère éviter de prendre parti xen faveur de
telle ou telle interprétation douteuse, et décide de laisser la dixfficulté de

côté en tranchant le différend sur la base d’autres considérations juridi -
quement pertinentes et suffisantes. C’est là la marque d’une saine pru -
dence judiciaire. Mais encore faut-il qu’il soit juridiquement possible,
compte tenu des données particulières de l’affaire, de juger xsans prendre
parti sur le sens de la norme dont la portée est douteuse. Ce n’esxt pas

toujours le cas. Et, par exemple, ce n’est pas le cas dans la présxente affaire,
pour les raisons que j’ai exposées plus haut : le traité de 1928, l’uti possi ‑
detis et les effectivités ne sont pas des bases juridiques alternatives,x placées
sur un pied d’égalité, entre lesquelles la Cour pouvait choisirx afin de tran -
cher la question de la souveraineté. Il fallait d’abord déciderx des effets du

traité de 1928 sur la souveraineté avant de pouvoir — le cas échéant —
examiner le reste. Décider ne peut pas se limiter à constater que xla tâche
est difficile: la Cour n’a pas rempli son office.
12. Certes, au moment où elle écrit, au paragraphe 56, qu’elle doit,
pour régler le différend, examiner les arguments des Parties autxres que

ceux fondés sur le traité, dont le sens lui paraît douteux, la Cour sait déjà
qu’elle trouvera dans l’examen des effectivités suffisammentx d’éléments
solides et peu contestables pour fonder une conclusion favorable à l’xattri -
bution de la souveraineté à l’une des Parties.
Mais cela ne change rien au problème. Pour les raisons déjà dites, la

Cour n’avait pas le loisir de choisir entre le traité et les effxectivités celui
des deux terrains qui lui paraissait le plus solide.

D’ailleurs, si les Parties avaient plaidé uniquement sur la base dxu traité,
la Cour n’aurait certainement pas éludé son devoir d’interprétation, dont

l’accomplissement peut être difficile, mais n’est jamais imposxsible.
13. J’ajouterai, pour prévenir une possible objection, que le devoir du juge
d’interpréter un traité invoqué par une partie, lorsqu’ilx n’existe pas de possi -
bilité juridique de se fonder sur une base strictement alternative, nxe se limite
pas à l’hypothèse dans laquelle la stipulation invoquée a poxur objet de définir

une règle de caractère général et impersonnel, une norme àx proprement par -
ler, c’est-à-dire abstraite et permanente. Le devoir d’interpréter vaut tout aussi
bien dans le cas où, comme en l’espèce, la clause litigieuse coxnfère un titre
spécifique à une partie, notamment un titre de souveraineté. xIl n’y a aucune
raison de déroger, en pareil cas, au principe fondamental selon lequexl le juge

n’a pas le droit de tirer argument de l’obscurité du traité xpour s’abstenir de
l’interpréter. Je regrette que la Cour ait méconnu ce principe xen l’espèce.
14. Cela dit, je pense que la Cour serait parvenue au même résultat
final si elle avait procédé comme elle aurait dû le faire.
15. Elle aurait d’abord constaté que, sauf à priver de tout effetx le der -
er
nier membre de phrase du premier alinéa de l’article 1 du traité de 1928,

113

6 CIJ1034.indb 222 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 733

adopt; that is not unusual. But it is the court’s duty to decide, irrespecxtive
of its doubts — doubts which it is moreover perfectly entitled to express

at the very moment when it does decide.
11. Admittedly, there are cases when, faced with a relatively obscurely
worded norm, the court prefers to avoid coming down in favour of one
particular questionable interpretation, and decides to set the difficulxty
aside and settle the dispute on the basis of other legally relevant and xsuf -

ficient considerations. That is the mark of healthy judicial caution. xHow-
ever, it still has to be legally possible, given the particular facts ofx the
case, to rule without establishing the meaning of the norm whose scope ixs
in doubt. That is not always the case. For example, it is not the case ixn
this instance, for the reasons which I have set out above : the 1928 Treaty,
the uti possidetis and the effectivités are not alternative legal bases, which

are on an equal footing, and between which the Court could choose in
order to settle the issue of sovereignty. It was necessary first to dextermine
the effects of the 1928 Treaty on sovereignty before the rest could — if
appropriate — be examined. Deciding cannot mean merely noting that
the task is difficult : the Court has not done its duty.

12. Admittedly, when it writes, in paragraph 56, that, in order to
resolve the dispute, it must examine the arguments of the Parties which x
are not based on the Treaty, whose meaning it regards as being in doubt,x
the Court already knows that when it considers the effectivités it will find

sufficient robust and relatively uncontentious evidence on which to basxe a
conclusion in favour of awarding sovereignty to one of the Parties.
However, that does not alter the problem. For the reasons which I have
already stated, the Court was not at leisure to choose between the Treatxy
and the effectivités on the basis of which of the two grounds appeared to
be the more robust.

Moreover, if the Parties had pleaded solely on the basis of the Treaty, x
the Court would certainly not have evaded its duty of interpretation, thxe
performance of which may be difficult but is never impossible.
13. I would add, to anticipate a possible objection, that a court’s duty
to interpret a treaty which has been adduced by a party, when it is not

legally possible to rely on a strictly alternative basis, is not limitedx to cases
in which the provision invoked seeks to define a rule of a general andx
impersonal nature, a genuine norm, that is to say, one which is abstract
and permanent. The duty to interpret is equally applicable in those casexs,
like the present one, in which the contentious clause confers a specific title

on a party, notably a title to sovereignty. In such cases there is no rexason
to derogate from the fundamental principle that a court is not entitled to
cite the obscurity of the treaty as justification for not interpretingx it. I
regret that the Court disregarded that principle in this case.
14. Having said that, I think that the Court’s final conclusion would
have been the same if it had proceeded as it ought to have done.

15. It would first have noted that, unless the last clause of the first xpara
graph of Article I of the 1928 Treaty were rendered ineffective, it must

113

6 CIJ1034.indb 223 7/01/14 12:43 734 différend territoriaxl et maritime (op. ind. axbraham)

il faut nécessairement admettre que certaines au moins des formationsx en

litige dans la présente phase de la procédure appartiennent à lxa Colombie
en vertu du traité, comme faisant partie de « l’archipel de San Andrés ».
En effet, cette disposition implique que des îles autres que San Andrés,
Providencia et Santa Catalina font partie de l’« archipel de San Andrés»
au sens du traité, et ces autres îles ne peuvent être que celles qui sont pré -

sentement en litige, ou certaines d’entre elles au moins. La thèsex nicara -
guayenne selon laquelle « l’archipel comprend uniquement les îles de
San Andrés, Providencia et Santa Catalina » (arrêt, par. 48) est incompa -
tible avec le traité, dès lors qu’elle le vide de son sens. Un simple coup
d’œil sur la carte permet de conclure que l’archipel comprend axu moins
e
— une fois que l’on a écarté toutes les îles situées àx l’ouest du 82 méri-
dien, que le protocole de 1930 déclare ne pas appartenir à l’arxchipel au
sens du traité — les cayes d’Alburquerque et de l’Est-Sud-Est, qui
sont les plus proches de San Andrés. Il est donc certain que ces îles sont
colombiennes en vertu du traité, et la Cour aurait dû le constaterx, au

lieu d’indiquer prudemment qu’elles « pourraient, vu leur situation géo -
graphique …, être considérées comme faisant partie de l’archipel » (ibid.,
par. 53), avant d’ajouter que ce critère géographique n’étaitx pas décisif.
16. A mon avis, il y aurait suffisamment d’éléments pour considéxrer
que les îles de Roncador, Quitasueño et Serrana étaient aussi rxegardées,

en 1928, comme faisant partie de l’archipel de San Andrés, mais il n’est
pas nécessaire de trancher cette question, car ces trois formations sxont
expressément exclues de l’attribution de la souveraineté à lxa Colombie par
le second alinéa de l’article 1erdu traité. Le fait qu’elles le soient au motif
que leur possession faisait à l’époque l’objet d’un litigxe entre la Colombie

et les Etats-Unis d’Amérique, litige qui a par la suite disparu du fait que
les Etats-Unis ont renoncé à leur revendication, ne change rien au constat
indiscutable que le traité de 1928 ne confère pas par lui-même un titre de
souveraineté à la Colombie sur les trois formations en cause. Dèxs lors, la
Cour pouvait laisser non résolue la question de savoir si Roncador, Qxui -

tasueño et Serrana faisaient partie de l’archipel de San Andrés au sens où
les deux Etats comprenaient cette notion en 1928.
17. Enfin, il me semble que Bajo Nuevo et Serranilla sont des îles trop
éloignées de San Andrés pour être raisonnablement considérées, à pre -
mière vue, comme faisant partie de l’archipel, et qu’une telle présomption

doit être retenue à moins que des éléments suffisamment proxbants tirés des
travaux préparatoires du traité de 1928 ne démontrent le contraxire. Or, la
Colombie n’a pas fourni de tels éléments au soutien de sa revenxdication.
18. Je conclus donc que le traité de 1928 confère à la Colombie la xsou -
veraineté non seulement sur San Andrés, Providencia et Santa Catalina

(ces trois îles n’étant plus en cause dans la présente phasxe de la procé -
dure), mais aussi sur les cayes d’Alburquerque et de l’Est-Sud-Est ; mais
qu’en revanche il ne confère la souveraineté à aucune des dexux Parties sur
les autres formations maritimes en litige.
19. Pour ces dernières — Quitasueño, Serrana, Roncador, Serranilla,

Bajo Nuevo — mais seulement pour elles, la Cour devait passer à l’exa -

114

6 CIJ1034.indb 224 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 734

inevitably be acknowledged that at least some of the features in disputex in
the present phase of the proceedings belong to Colombia on the basis of

the Treaty, since they form part of the “San Andrés Archipelago”x. That
provision in fact implies that islands other than San Andrés, Providencia
and Santa Catalina form part of the “San Andrés Archipelago” under the
Treaty, and those other islands can be none other than those which are
presently in dispute, or certain of them at least. Nicaragua’s positixon, that

“the Archipelago comprises only the islands of San Andrés, Providexncia
and Santa Catalina” (para. 48 of the Judgment) is incompatible with the
Treaty, since it renders it meaningless. A simple glance at the map is sxuf -
ficient to conclude — once you disregard all the islands to the west of the
82º W meridian, which the 1930 Protocol declares not to belong to the
Archipelago under the Treaty — that the Archipelago includes at least the

Alburquerque Cays and the East-Southeast Cays, which lie closest to
San Andrés. Those islands therefore definitely belong to Colombia underx
the Treaty, and the Court ought to have noted that fact, instead of cau -
tiously indicating that “given their geographical location” they “xcould be
seen as forming part of the Archipelago” (ibid., para. 53), before adding

that this geographical criterion was not decisive.
16. In my opinion, there is sufficient evidence to consider that in 1928
the islands of Roncador, Quitasueño and Serrana were also regarded as
forming part of the San Andrés Archipelago, but it is not necessary to
settle that question, since the second paragraph of Article I of the Treaty

expressly precludes sovereignty over those three features from being
attributed to Colombia. The fact that the reason given is that their appxur -
tenance was in dispute between Colombia and the United States of Amer-
ica at the time, a dispute which subsequently disappeared when the Unitexd
States renounced its claim, does not alter the indisputable fact that thxe
1928 Treaty does not in itself confer a title of sovereignty on Colombia

over the three features in question. The Court was therefore able to leaxve
the issue unresolved of whether Roncador, Quitasueño and Serrana
formed part of the San Andrés Archipelago in the sense in which the two
States understood that notion in 1928.
17. Finally, it seems to me that Bajo Nuevo and Serranilla are too far

away from San Andrés to be reasonably regarded, at first sight, as form -
ing part of the Archipelago, and that this assumption must be made,
unless there is sufficiently convincing evidence to the contrary in thex
travaux préparatoires of the 1928 Treaty. However, Colombia did not
provide any such evidence in support of its claim.

18. I therefore conclude that the 1928 Treaty accords Colombia sover -
eignty not only over San Andrés, Providencia and Santa Catalina (sinxce
these three islands are no longer at issue in the present phase of the pro -
ceedings), but also over the Alburquerque Cays and the East-Southeast
Cays; however, it does not accord either of the two Parties sovereignty
over the other maritime features in dispute.

19. In respect of the latter — Quitasueño, Serrana, Roncador, Serra -
nilla and Bajo Nuevo — but of them alone, the Court had to move to the

114

6 CIJ1034.indb 225 7/01/14 12:43 735 différend territoriaxl et maritime (op. ind. axbraham)

men des arguments fondés sur l’uti possidetis et sur les effectivités post-
coloniales. A cet égard, j’adhère à la suite du raisonnementx de l’arrêt : le

principe de l’uti possidetis ne permet pas d’établir un titre en faveur de
l’une ou l’autre des Parties, les effectivités sont en faveurx de la Colombie.

20. En définitive, mes sérieuses réserves sur le raisonnement suixvi par
l’arrêt ne m’ont pas empêché de voter en faveur du point 1 du dispositif,

car ma conclusion rejoint celle de mes collègues.

* * *

II. Le tracé d’une ligne médxiane provisoire
en tant que première éxtape de la méthode de fxixation
de la frontière maritxime

21. En ce qui concerne la délimitation maritime, mon désaccord porte

moins sur ce qu’a fait la Cour — j’approuve d’ailleurs le résultat final de
l’opération, et j’ai voté en faveur des points 4 et 5 du dispositif — que sur
la manière de le présenter, qui m’apparaît assez largement fxallacieuse. En
résumé, mon opinion est que, si la Cour affirme qu’elle suit la méthode
traditionnelle telle qu’elle a été exposée, notamment, dans xl’arrêt rendu en

l’affaire Roumanie c. Ukraine (Délimitation maritime en mer Noire (Rou ‑
manie c. Ukraine), arrêt, C.I.J. Recueil 2009, p. 61), elle s’en écarte en
réalité très largement, et en vérité elle ne peut pas faire autrement car il est
manifeste que cette méthode est inadaptée à la présente affxaire.

22. La méthode en question est rappelée aux paragraphes 190 à 193 de

l’arrêt. Elle consiste à tracer d’abord une ligne médianex provisoire,
c’est-à-dire une ligne située à égale distance des côtes opposéesx des deux
Etats et qui génèrent les droits à des espaces maritimes qui sex che -
vauchent — ce chevauchement étant la raison même pour laquelle il y a
lieu de procéder à une délimitation. Lorsque les côtes pertinentes sont

adjacentes, la ligne provisoire est appelée ligne d’équidistancxe, mais cela
ne change rien en substance et ce n’est d’ailleurs pas le cas de lx’espèce. La
méthode consiste, dans un deuxième temps, à ajuster ou à déxplacer la ligne
provisoire ainsi obtenue pour tenir compte des circonstances particulièxres
qui, s’il en existe, appelleraient un tel ajustement ou déplacemenxt en vue

de parvenir à une solution équitable. Enfin, dans une troisièxme étape, la
Cour doit vérifier que la délimitation obtenue au terme de l’xétape précé -
dente n’aboutit pas à attribuer aux Parties des espaces maritimes xdont le
rapport de superficie serait hors de proportion, de façon marquéxe, avec la
longueur de leurs côtes pertinentes respectives — les côtes qui génèrent les
droits à des espaces qui se chevauchent.

23. La Cour examine les arguments par lesquels le Nicaragua a cher -
ché à la convaincre que ladite méthode était inappropriéex au cas d’espèce,

115

6 CIJ1034.indb 226 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 735

examination of the arguments based on the uti possidetis and the
post-colonial effectivités. In this regard, I support the Judgment’s subse -

quent reasoning: a title in favour of one or other of the Parties cannot be
established on the basis of the principle of uti possidetis ; the effectivités
are in Colombia’s favour.
20. Ultimately, my serious reservations about the reasoning in the
Judgment did not prevent me from voting in favour of point 1 of the

operative part, since my conclusion is the same as that of my colleaguesx.

* * *

II. The Construction of a Prxovisional Median Line
as the First Stage in thex Method for Fixing
the Maritime Boundary

21. As far as the maritime delimitation is concerned, my disagreement

relates less to what the Court has done — moreover, I agree with the end
result of the process, and I voted in favour of points 4 and 5 of the oper -
ative part — than to how it is presented, which appears to me to be
largely fallacious. In short, my opinion is that, although the Court staxtes
that it is following the traditional method, as described in particular xin its

Judgment in the case between Romania and Ukraine (Maritime Delimita ‑
tion in the Black Sea (Romania v. Ukraine), Judgment, I.C.J. Reports 2009,
p. 61), in reality it diverges very considerably from it and actually it can -
not do otherwise, since it is clear that the said method is inappropriatxe in
the present case.
22. The method in question is recalled in paragraphs 190 to 193 of the

Judgment. It consists of first constructing a provisional median line,x that
is to say, a line which is at equal distance from the opposite coasts ofx the
two States which generate entitlements to overlapping maritime spaces —
those overlapping entitlements being the very reason why it is necessaryx
to effect a delimitation. Where the relevant coasts are adjacent, the provi -

sional line is termed an equidistance line, but that does not make any
substantial difference and moreover is not the case here. The second sxtage
is to adjust or shift the provisional line thus obtained in order to take
account of any particular circumstances which might require the line to x
be adjusted or shifted in order to achieve an equitable solution. Finallxy,

in a third stage, the Court must check that the maritime areas awarded txo
the Parties by virtue of the delimitation obtained at the end of the previ -
ous stage are not markedly disproportionate to their respective relevantx
coasts — the coasts which generate the entitlements to the overlapping
spaces.

23. The Court considers the arguments by which Nicaragua sought to
convince it that the said method was inappropriate in the present case oxn

115

6 CIJ1034.indb 227 7/01/14 12:43 736 différend territoriaxl et maritime (op. ind. axbraham)

parce qu’il n’était pas opportun, compte tenu de la situation gxéographique

particulière, de commencer par le tracé d’une ligne médiane xprovisoire.
Elle reconnaît que la « démarche en trois temps ne doit pas être appliquée
de façon mécanique », et « qu’il ne serait peut-être pas toujours opportun
de commencer par l’établissement d’une ligne d’équidistanxce (ou médiane)
provisoire» (arrêt, par. 194). Mais elle écarte les arguments du Nicara -

gua, et affirme que, s’il existe sans doute des circonstances particxulières
justifiant d’ajuster la ligne médiane provisoire, il n’existe aucune raison de
ne pas commencer par tracer celle-ci et de ne pas l’utiliser comme point de
départ du processus de délimitation. En conséquence, la Cour affixrme
qu’elle s’en tiendra à sa « méthode de référence » (ibid., par. 199), et elle

procède ainsi — ou plutôt elle prétend procéder ainsi — aux para -
graphes 200 à 204 (première étape : construction de la ligne médiane pro -
visoire), aux paragraphes 205 à 238 (deuxième étape : ajustement ou
déplacement de la ligne provisoire), et aux paragraphes 239 à 247 (troi -
sième étape: vérification de l’absence de disproportion).

24. Pourtant, il est manifeste que la construction d’une ligne médianex
provisoire comme point de départ de la délimitation est non seulemxent
hautement inopportune en l’espèce, mais qu’elle est même quaxsi impos -
sible.
25. La raison en est très simple. Le chevauchement des droits qui rend

nécessaire la délimitation en l’espèce ne provient pas de cex que deux côtes
opposées (ou adjacentes) génèrent des projections qui se chevxauchent dans
une zone intermédiaire, comme c’est habituellement le cas. Ici, lex chevau -
chement des droits provient de ce que, à l’intérieur de la zonex économique
exclusive mesurée à partir des côtes nicaraguayennes se trouvenxt des îles

appartenant à la Colombie et qui génèrent, dans toutes les directions, un
droit à une zone économique exclusive au profit de ce dernier Etxat. Autre -
ment dit, le chevauchement ne se produit pas seulement entre les côtexs
nicaraguayennes et les îles colombiennes (c’est-à-dire dans la partie située
à l’ouest des îles colombiennes et à l’est de la côte xnicaraguayenne) ; il se

produit aussi dans les espaces situés au nord, à l’est et au suxd des îles
colombiennes — et même entre celles-ci. C’est ce que fait très clairement
apparaître le croquis n o7 figurant dans l’arrêt (p. 687), qui représente la
«zone maritime pertinente », c’est-à-dire la zone de chevauchement des
droits à l’intérieur de laquelle la Cour est appelée à opxérer la délimitation.

26. Dès lors, il me paraît clair qu’aucune « ligne médiane » ne peut
rendre compte de la réalité géographique qui était soumise àx l’examen de
la Cour, non pas en raison de telle ou telle « circonstance particulière per -
tinente» qui justifierait l’ajustement d’une ligne provisoire sans en rendre
le tracé impossible, mais en raison des données essentielles du dixfférend

soumis à la Cour, qui rendent la notion même de « ligne médiane »
dépourvue de signification en l’espèce.
27. Il est certes possible de tracer une ligne située à égale distaxnce des
côtes du Nicaragua et des côtes occidentales des îles colombienxnes, et c’est
ce que fait la Cour, en affirmant que, ce faisant, elle a satisfait àx la pre -

mière étape de sa « méthode de référence». Mais il suffit de jeter un coup

116

6 CIJ1034.indb 228 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 736

the grounds that the particular geographical situation was one in which x
the Court should not begin by constructing a provisional median line. Itx

acknowledges that the “three-stage process is not . . . to be applied in a
mechanical fashion” and “that it will not be appropriate in every xcase to
begin with a provisional equidistance/median line” (Judgment, para. 194).
However, it dismisses Nicaragua’s arguments and states that, althoughx
there are undoubtedly particular circumstances which justify adjusting

the provisional median line, there is no reason not to begin by construcxt -
ing such a line nor to use it as a starting-point for the delimitation. xCon-
sequently, the Court affirms that it will adhere to its “standard mexthod”
(ibid., para. 199), and it proceeds to do so — or rather it claims to pro -
ceed to do so — in paragraphs 200 to 204 (first stage : construction of the
provisional median line), in paragraphs 205 to 238 (second stage : adjust -

ment or shifting of the provisional line), and in paragraphs 239 to 247
(third stage: disproportionality test).

24. Nevertheless, it is obvious that the construction of a provisional
median line as a starting-point for the delimitation is not only highly

inappropriate in this case, but that it is even virtually impossible.

25. The reason for this is very simple. The overlapping entitlements
which make the delimitation necessary in this instance do not exist
because two opposite (or adjacent) coasts are generating projections

which overlap in an intermediate area, as is usually the case. Here, thex
overlapping entitlements occur because, within the exclusive economic
zone measured from the Nicaraguan coast, there are islands belonging to x
Colombia which generate an entitlement to an exclusive economic zone
for that State in all directions. In other words, the overlapping does nxot
only occur between the Nicaraguan coast and the Colombian islands

(that is to say, in the area to the west of the Colombian islands and to the
east of the Nicaraguan coast); it also occurs in the areas to the north, east
and south of the Colombian islands — and even between them. This is
shown very clearly on sketch-map No. 7 in the Judgment (p. 687), which
depicts the “relevant maritime area”, that is to say, the area of xoverlap -

ping entitlements within which the Court is called upon to effect the x
delimitation.
26. Plainly, therefore, no “median line” can take account of the geo -
graphical reality which was submitted for the Court’s consideration, xnot
because of any “relevant particular circumstance” which would justxify the

adjustment of a provisional line without making it impossible to con -
struct it in the first place, but because of the essential facts of thxe dispute
brought before the Court, which make the very notion of a “median linxe”
meaningless in the present case.
27. It is admittedly possible to construct a line which is equidistant
from the Nicaraguan coast and the west-facing coasts of the Colombian

islands, and that is what the Court does, affirming that in so doing itx has
completed the first stage of its “standard method”. But a glancex at that

116

6 CIJ1034.indb 229 7/01/14 12:43 737 différend territoriaxl et maritime (op. ind. axbraham)

d’œil à cette ligne, qui est représentée au croquis n 8 (p. 701), pour s’aper-

cevoir qu’elle n’a de «médiane» que le nom: elle est certes à égale distance
des côtes nicaraguayennes (plus exactement des îles nicaraguayennes
adjacentes à la côte continentale de cet Etat), sur son flanc oxuest, et des
côtes occidentales des îles colombiennes, qui se situent aussi loixn à l’est de
la ligne en cause. Mais elle ne tient aucun compte — et elle ne peut par

construction tenir aucun compte — de tout l’espace situé à l’est des îles
colombiennes, qui fait pourtant partie lui aussi de la zone de chevauchex -
ment. Il ne s’agit pas là d’une « circonstance particulière» qui justifierait
un ajustement ou un déplacement postérieur. Il s’agit d’un dxéfaut rédhibi -
toire qui prive la ligne de son caractère supposé «médian». Cela s’explique

par un élément propre à l’espèce : la Cour n’a pu tracer cette ligne qu’en
prenant des points de base, en ce qui concerne la partie colombienne,
situés exclusivement sur les côtes occidentales des îles appartxenant à cet
Etat. Elle n’a pas pu retenir de point de base sur les parties orientxales,
septentrionales et méridionales des côtes de ces îles puisque cxes parties ne

sont pas tournées vers les côtes nicaraguayennes. Or, comme je l’xai rappelé
plus haut, c’est l’ensemble des côtes insulaires de la Colombiex, et non pas
seulement la partie occidentale de ces côtes, qui génèrent les droits à une
zone économique exclusive qui chevauchent ceux du Nicaragua.
28. En d’autres termes, pour pouvoir tracer une ligne qui ait au moins

l’apparence d’une « ligne médiane» — mais selon moi c’est à peine si elle
en a l’apparence —, la Cour a dû délibérément négliger la plus grande
partie des côtes pertinentes de la Colombie. Or, une ligne médianex, pour
remplir la fonction qui lui est assignée dans le processus de délixmitation,
doit prendre en compte l’ensemble des « côtes pertinentes » des Etats en

présence, c’est-à-dire l’ensemble des côtes qui génèrent les projections
créant les chevauchements de droits rendant nécessaire la délimxitation.
29. L’arrêt lui-même le rappelle dans son paragraphe 191 : la ligne
médiane doit être « tracée à partir des points les plus pertinents des côtes
des deux Etats concernés ». Il s’agit certes de points choisis, mais ils ne

peuvent pas l’être de n’importe quelle manière : pour qu’ils soient « les
plus pertinents », il faut qu’ils rendent compte de façon satisfaisante de
l’ensemble des « côtes pertinentes » et non pas d’une partie de ces côtes
seulement. Or, pour ce qui est des îles colombiennes, l’arrêt ixndique à
juste titre que les côtes pertinentes sont constituées par « l’intégralité de la

côte de ces formations, et non leurs seules côtes occidentales » (par. 151).
D’où il résulte qu’une ligne médiane répondant à lax définition de la
«méthode de référence » devrait être tracée à partir de points de base
situés sur toutes les parties des côtes des îles colombiennes, xet non pas
seulement sur leurs parties occidentales. Mais ce n’est évidemmentx pas

possible en l’espèce.
30. Au lieu d’en déduire que la construction d’une ligne médianex — fût-
elle provisoire — est pour le moins inappropriée, pour ne pas dire im-
possible, en l’espèce, la Cour décide tout de même d’en tracer une, en
s’abstenant de prendre en compte (simplement parce qu’elle ne le xpeut

pas) la plus grande partie des côtes des îles colombiennes. Ce faxisant, elle

117

6 CIJ1034.indb 230 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 737

line, which is shown on sketch-map No. 8 (p. 701), is sufficient to realize
that it is “median” in name only : it may be equidistant from the Nicara -

guan coast (more precisely from the Nicaraguan islands adjacent to that
State’s mainland coast), on its western side, and the western coasts of the
Colombian islands, on its eastern side. However, it does not take any
account — indeed the manner of its construction means that it cannot
take any account — of the entire area to the east of the Colombian

islands, which nonetheless also forms part of the overlapping area. Thisx is
not a “particular circumstance” which would justify a subsequent axdjust -
ment or shifting of the line. It is a fundamental defect which deprives xthe
line of its alleged “median” character. This can be explained by ax specific
characteristic of the case : the Court could only construct that line by
taking base points, as far as Colombia is concerned, which were located x

exclusively on the west-facing coasts of the islands belonging to
that State. It could not adopt any base points on the east-, north- and
south-facing coasts of those islands since they do not face the Nicaraguan
coast. However, as I recalled above, all of the coasts of Colombia’s x
islands, not just the west-facing parts of those coasts, generate

entitlements to an exclusive economic zone which overlap with those of
Nicaragua.
28. In other words, in order to be able to construct a line which has at
least the semblance of a “median line” — although in my view even that
is debatable — the Court deliberately had to ignore the majority of

Colombia’s relevant coasts. However, in order to perform its designated
function in the delimitation process, a median line must take into accouxnt
all the “relevant coasts” of the States present, that is to say, axll the coasts
which generate the projections creating the overlapping entitlements
which make the delimitation necessary.
29. The Judgment itself recalls that point in paragraph 191: the median

line has to be “constructed using the most appropriate base points onx the
coasts of the Parties”. These points are admittedly chosen, but they xcan -
not be chosen in just any way : in order for them to be “the most appro -
priate”, they must take satisfactory account of all the “relevant xcoasts”
and not just one part of those coasts. However, as far as the Colombian x

islands are concerned, the Judgment rightly points out that the relevantx
coasts constitute “the entire coastline of these islands, not merely xthe
west-facing coasts” (para. 151). This suggests that a median line corre -
sponding to the definition in the “standard method” would have to be
constructed from base points on all the coasts of the Colombian islands,

and not only on their west-facing parts. Clearly, however, that is not pxos -
sible in this case.

30. Instead of concluding from this that the construction of a median
line — albeit a provisional one — is at the very least inappropriate, if not
impossible, in this case, the Court decides to construct one all the samxe

without taking into account (simply because it cannot) the majority ofx the
coasts of the Colombian islands. In so doing, it appears to forget in paxra -

117

6 CIJ1034.indb 231 7/01/14 12:43 738 différend territoriaxl et maritime (op. ind. axbraham)

paraît oublier aux paragraphes 200 à 204, dans lesquels elle retient des

points de base situés exclusivement, du côté colombien, sur la xfaçade oc-
cidentale des îles, ce qu’elle a expliqué aux paragraphes 151 et 191.
31. Cela lui permet, certes, de tracer une ligne (représentée au croqxuis
no 8). Mais cette ligne n’est « médiane» qu’en ce qui concerne l’une des
parties de la « zone pertinente » à délimiter (la zone qui figure au croquis
o
n 7); elle est sans signification aucune pour le reste. Selon moi, donc, la
ligne tracée ne saurait se voir conférer la valeur qui est celle d’une « ligne
médiane», à savoir celle d’un point de départ acceptable de la délimita -
tion, dont on ne s’écartera ensuite que pour procéder à des xajustements
ou à des déplacements par nature limités, destinés à tenixr compte de cir -

constances particulières.
32. L’arrêt en fait d’ailleurs l’aveu implicite, dans la suite dxe sa motiva -
tion, de deux manières.
En premier lieu, après avoir ajusté la ligne provisoire en la déxplaçant
considérablement vers l’est (pour la rapprocher, donc, des îlexs colom -

biennes), la Cour constate que, même après cet ajustement, le réxsultat
obtenu ne serait pas équitable si la ligne « pénétrait dans des secteurs de la
zone pertinente situés, vers le nord, au-delà du point 1 et, vers le sud,
au-delà du point 5 », c’est-à-dire au nord et au sud des principales îles
colombiennes, et qu’en outre la ligne en cause priverait le Nicaraguax des

espaces situés à l’est de ces îles, espaces « dans lesquels se projette sa côte
continentale» (arrêt, par. 236). C’est parfaitement exact, mais n’est-ce pas
une façon de reconnaître que la ligne provisoire n’est pas aptex à remplir
la fonction qui devrait être la sienne, pour une grande partie de la xzone
dans laquelle la délimitation est à opérer, à savoir tous lexs secteurs situés

au nord, au sud et à l’est des principales îles colombiennes ?
En second lieu, et en conséquence de ce qui précède, la Cour esxt
conduite à construire deux lignes horizontales suivant des parallèxles pas -
sant au nord par le point 1 (situé au nord de Santa Catalina, et à peu près
à la hauteur de Roncador) et au sud par le point 9 (situé à la hauteur des

cayes de l’Est-Sud-Est), et ce en vue de délimiter la zone située à l’est des x
îles colombiennes (ibid., par. 237). Or, il est difficile de voir dans ces deux
lignes horizontales un simple « ajustement» ni même un « déplacement»
de la ligne provisoire. Il s’agit plutôt du tracé de lignes n’xayant rien à
voir — sauf, pour la première, son point de départ — avec la ligne provi -

soire. Il en va de même de l’ajout de pas moins de quatre poonts dxe fron -
tière maritime (les points 6 à 9 figurant sur le croquis n 11, p. 714) dans
la partie méridionale de la zone à délimiter, qui, plutôt quxe d’ajuster ou
de déplacer la ligne provisoire, se chargent en réalité de la cxompléter.

33. En somme, après avoir qualifié de « ligne médiane » une ligne ne
méritant pas vraiment cette qualification, la Cour appelle « ajustement»
ou «déplacement» une opération qui ne mérite pas vraiment d’être appe -
lée ainsi. A moins que cette opération (le tracé des deux lignxes horizon -
tales et la fixation des points 6 à 9) ne soit conçue par l’arrêt comme une

étape postérieure à l’ajustement ou au déplacement eux-mêmes. Mais

118

6 CIJ1034.indb 232 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 738

graphs 200 to 204, in which it selects base points which, on the Colom -
bian side, are located exclusively on the western sides of the islands, xwhat

it explained in paragraphs 151 and 191.
31. It is true that this enables it to construct a line (depicted on
sketch-map No. 8). But that line is only “median” with respect to one
part of the “relevant area” to delimit (the area shown on sketch-map
No. 7); it is otherwise entirely meaningless. In my view, therefore, the line x

constructed cannot be regarded as a “median line”, that is, as an xaccept -
able starting-point for the delimitation, which will subsequently only be
adjusted or shifted to a necessarily limited extent, in order to take acxcount
of particular circumstances.

32. Moreover, further on in its reasoning the Judgment implicitly

acknowledges that fact, in two ways.
First, after adjusting the provisional line by shifting it considerably x
eastwards (in order, therefore, to move it closer to the Colombian islaxnds),
the Court notes that even after that adjustment the result would not be x
equitable if the line “extend[ed]... into the parts of the relevant axrea north

of point 1 or south of point 5”, that is to say, to the north and south of
the principal Colombian islands, and that furthermore the line in ques -
tion would cut off Nicaragua from the areas to the east of those islanxds,
areas “into which the Nicaraguan coast projects” (Judgment, para.x 236).
That is perfectly true, but does it not constitute an acknowledgment thaxt

the provisional line is not fit for purpose, with regard to a large paxrt of
the area in which the delimitation is to be effected, that is to say, xall the
sectors to the north, south and east of the principal Colombian islands ?

Second, and as a consequence of the foregoing, the Court is induced to
construct two horizontal lines along lines of latitude passing to the noxrth

through point 1 (which is located to the north of Santa Catalina, and
approximately level with Roncador) and to the south through point 9
(which is located level with East-Southeast Cays) with a view to delimit -
ing the area to the east of the Colombian islands (ibid., para. 237). How -
ever, it is difficult to regard these two horizontal lines as a mere

“adjustment” or even “shifting” of the provisional line. With the excep -
tion of the starting-point of the first line, those lines are actually entirely
unrelated to the provisional line. The same goes for the addition of no x
fewer than four maritime frontier points (points 6 to 9 on sketch-map
No. 11, p. 714) in the southern part of the area to be delimited which,

rather than adjusting or shifting the provisional line, are in fact suppxle -
mentary to it.
33. In short, after describing as a “median line” a line which does noxt
really merit that description, the Court terms an “adjustment” or x“shift”
a process which does not really merit being termed as such. Perhaps the x
Judgment envisaged that process (the construction of two horizontal linxes

and the fixing of points 6 to 9) as a separate stage after the adjustment or
shifting of the line. But if so, would that not be adding another stage —

118

6 CIJ1034.indb 233 7/01/14 12:43 739 différend territoriaxl et maritime (op. ind. axbraham)

n’est-ce pas alors ajouter une étape supplémentaire — et d’une impor -

tance décisive en l’espèce — à la « méthode traditionnelle » (ou « de réfé-
rence») que la Cour a pourtant promis de suivre au paragraphe 199 ?
34. Je ne veux pas dire que la Cour ait eu tort de délimiter les espaces x
constituant la zone pertinente comme elle l’a fait. Je pense au contrxaire
qu’elle a adopté la solution la plus raisonnable, et que chaque éxtape de sa

construction était intrinsèquement justifiée. Mais mon opinioxn est qu’il
eût été plus clair et plus franc de la part de la Cour de reconxnaître qu’elle
ne pouvait pas suivre en l’espèce sa méthode dite « de référence » parce
que le cadre géographique ne se prêtait absolument pas à l’axpplication de
cette méthode. Ainsi se trouvait-elle en l’espèce dans le cas où « des rai-

sons impérieuses propres au cas d’espèce » ne permettaient pas de tracer
la ligne médiane provisoire (Délimitation maritime en mer Noire (Rouma ‑
nie c. Ukraine), arrêt, C.I.J. Recueil 2009, p. 101, par. 116), ou, à tout le
moins, dans celui où l’application de la méthode de l’équxidistance était
« inappropriée » (Différend territorial et maritime entre le Nicaragua et le

Honduras dans la mer des Caraïbes (Nicaragua c. Honduras), arrêt,
C.I.J. Recueil 2007 (II), p. 741, par. 272, mentionné au paragraphe 194
du présent arrêt).
35. Je comprends que la Cour veuille donner à tous ceux qui l’ob -
servent, et d’abord aux Etats, le sentiment qu’elle ne procède xpas de

manière arbitraire pour parvenir à une solution équitable, maisx qu’elle
met en œuvre des techniques éprouvées et constantes. Et il est xparfaite -
ment vrai qu’il n’y a aucun arbitraire dans la démarche de la Cxour, mais
seulement la recherche scrupuleuse de la meilleure solution. Il est cepexn -
dant des affaires qui se présentent en des termes tellement spécxifiques qu’il

est, à tout prendre, préférable de reconnaître que la Cour dxoit s’écarter de
sa technique habituelle, en expliquant pourquoi, plutôt que de sacrifixer la
clarté et l’intelligibilité à l’apparence d’une illusoxire continuité.

(Signé) Ronny Abraham.

119

6 CIJ1034.indb 234 7/01/14 12:43 territorial and marixtime dispute (sep. op. abxraham) 739

and a decisively important one in this case — to the “traditional method”

(or “standard method”) to which the Court nevertheless promised xto
adhere in paragraph 199 ?
34. I do not wish to say that the Court was wrong to delimit the spaces
constituting the relevant area in the way that it did. On the contrary, xI
think that it adopted the most reasonable solution, and that each stage xin

its construction was intrinsically justified. However, my opinion is txhat it
would have been clearer and more honest of the Court to acknowledge
that it could not follow the so-called “standard” method in this cxase
because the geographical framework did not at all lend itself to the appli -
cation of that method. In this instance it thus found itself in the situxation

in which “compelling reasons . . . in the particular case” made it unfeasi -
ble to construct the provisional median line (Maritime Delimitation in the
Black Sea (Romania v. Ukraine), Judgment, I.C.J. Reports 2009, p. 101,
para. 116) or, at the very least, in one in which the application of the
equidistance method was “inappropriate” (Territorial and Maritime Dis ‑

pute between Nicaragua and Honduras in the Caribbean Sea (Nicaragua v.
Honduras), Judgment, I.C.J. Reports 2007 (II), p. 741, para. 272, men -
tioned in paragraph 194 of the present Judgment).
35. I understand that the Court wishes to give all its observers, and
first and foremost States, the impression that it does not use arbitraxry

methods to achieve an equitable solution, but that it implements proven x
and consistent techniques. And it is perfectly true that there is nothinxg
arbitrary about the Court’s approach, which is characterized merely bxy a
scrupulous search for the best solution. However, there are cases which x
are presented in such specific terms that it is, on the whole, preferaxble to

acknowledge that the Court needs to depart from its usual technique, and
to explain why, rather than to sacrifice clarity and intelligibility txo the
semblance of an illusory continuity.

(Signed) Ronny Abraham.

119

6 CIJ1034.indb 235 7/01/14 12:43

Document file FR
Document Long Title

Separate opinion of Judge Abraham

Links