How not to avoid using reference numbers as nouns in an academic article [closed]

I wrote a scientific article for a conference, and a reviewer criticized it as he said that in my manuscript “references are used as nouns”. I assume he was refering to instances of the following type, although he did not say so explicitly:

1. According to [20]
2. …is described in [17]
3. …as highlighted in [7]

(The conference proceedings specifies this in-text numerical reference style.)

Is it really wrong to use reference numbers in this way? In any case, how can I satisfy him by rewriting these examples to avoid doing this?

You are right in thinking the examples you cite are what the reviewer means by “references are used as nouns” (see IEEE Citation Reference, below).

Is this really wrong?

In my view it cannot be judged ‘wrong’ as at least one very respectable journal allows it — it is more a question of writing style and house style. First, the evidence that some journals allow it:

[IEEE (Institute of Electrical and Electronics Engineers) Citation Reference, 2009]

Why, then, does your reviewer not like it (and presumably the house style of your conference proceedings and other journals not allow it)? I think it is for the reason mentioned by @sumelic in his comment — a conviction that the sentence should be able to stand alone, without references.

Certainly the style you use is ugly if read aloud, and can be avoided because it won’t work with superscript reference numbers. To digress a little, there are three main citation styles for references (in my area of science), superscript numbers, numbers in parentheses written within the text, and author/date written within the text.

If one considers superscripts (which is probably the oldest form), you can write “DNA consists of a double helix¹”, but obviously not “According to ¹ DNA consists of a double helix”.

However the author/date style of referencing allows this. You can write “DNA consists of a double helix (Watson and Crick, 1953)” or “According to Watson and Crick (1953) DNA consists of a double helix”.

I imagine this latter is what has led to the use of the style of “According to [20]” with the later introduction of in-text numerals (easier to set in type). However some regard in-text numerals in the same way as superscripts, and disapprove of this usage.

How can it be replaced?

First check any ‘Instructions to Authors’ or past conference proceedings to find out what the house style is. If we consider your first example:

According to [20]

I suggest two alternatives.

(i) Refer to the authors of the paper:

According to Smith and Brown [20]

e.g.

(ii) Use the passive:

It has been suggested/proposed/reported [20]

e.g.

(In this case put the reference at the end of the sentence so that it cannot be considered part of it.)

I am sure you can adopt these for examples 2 and 3 yourself. Dealing with the individual quirks of reviewers is often galling, but one has to grit one’s teeth and do what is requested to get published.