The Invisible Constitution in Comparative Perspective

4.4.5.  Obviousness as a Requirement for Implications

There has been some disagreement as to whether, in order to be recognised 

as genuine, implications should have to satisfy a test of ‘obviousness’. Ryan 

Williams, for example, in discussing the Ninth Amendment, proposed such 

96 

Shirlaw v. Southern Foundries (1926) Ltd [1939] 2 KB 206, 227.

97 

Grice, Supra note 33–34.

---

134 

Jeffrey Goldsworthy

a test for constitutional implicatures,

98

 but is criticised by Larry Solum in this 

book for doing so.

99

  Solum points out that implicatures are not always obvi-

ous; sometimes ‘competent speakers would need to read carefully to see the 

implicature’.

100

 Some linguists regard implicatures as ‘always somewhat uncer-

tain’, because the communicative principles can conflict with one another, 

and because doubts can arise at any stage of the complex reasoning process 

that is involved.

101

 Other forms of contextual enrichment such as ellipses are 

also debatable. For example, s. 92 of the Australian Constitution is clearly 

elliptical, but there was much disagreement about how the ellipsis should be 

filled.

102

On the other hand, to show that utterance meanings are shaped partly by 

what I have just described as implicit assumptions, a test of obviousness may 

be appropriate. As I have previously argued:

implicit assumptions are unstated, or even unnoticed, precisely because they 

are obvious. If so, there can be no such thing as an implicit assumption that 

does not comply with the obviousness test, and so the test is necessary.

103

For other kinds of inexplicit content, including implicatures, ‘we do not 

require obviousness in everyday interpretation. We employ a process of rea-

soning in which speakers’ meanings are inferred from various kinds of evi-

dence, with varying degrees of strength, and we are willing to accept a higher 

risk of error than are the courts in interpreting legal documents’.

104

Obviousness, then, is not normally required. But it is open to courts to be 

more cautious. They may be especially wary of error because they are not 

generally permitted to rewrite the laws they are charged with interpreting. 

They might regard attributing a non-existent implication to a law to be worse 

than failing to identify a genuine one through excessive caution. That would, 

of course, be debatable, but they might consequently subject themselves to 

   98 

Williams, Supra note 42, 544.

   99 

L. Solum Section 

3.6.3

. See also H. Asgeirrson, ‘On the Possibility of Non-literal Legislative 

Speech’, in A. Capone and F. Poggi (eds.), Pragmatics and Law: Theoretical and Practical 

Perspectives (Springer Verlag, 2017), 11 who disagrees with A. Marmor on this point.

100 

Solum, Supra note 99.

101 

C. Potts, ‘Presupposition and Implicature’, in S. Lappin and C. Fox (eds.), The Handbook of 

Contemporary Semantic Theory, 2nd edn (Wiley Blackwell 2015), 168, 183.

102 

See note 31.

103 

Goldsworthy, chapter note, 170. Note also that Mason CJ in the ACTV case did not require 

‘textual’ implications to be ‘necessary’: see notes 120–122.

104 

Goldsworthy, chapter note, 170. I did not at the time mention implicatures in this regard.

---