Report: Analysis of 'Whom' Usage in American English (COCA)

This report investigates the usage of the word 'whom' in American English, comparing its frequency in spoken and written contexts. The study utilizes the Corpus of Contemporary American English (COCA) to analyze data from 1991 to 2006, dividing the data into spoken and written segments. A hypothesis is tested, positing that 'whom' is used more frequently in written American English due to the greater emphasis on correct grammar. The analysis involves sentence parsing using the Natural Language ToolKit (NLTK) library to identify instances of 'whom' as the object of a verb or preposition. The results reveal that 'whom' appears more often in written English (2665 instances per million sentences) compared to spoken English (953 instances per million sentences), thus supporting the hypothesis. The report concludes that the prescriptive rule regarding 'whom' is more consistently followed in written American English, where grammatical correctness is prioritized. The report provides a detailed breakdown of the data and statistical analysis to support its findings.

Running head: GRAMMAR
GRAMMAR
Name of the Student
Name of the University
Author note

1GRAMMAR
Introduction
Often, non-native English speakers as well as native speakers, sometimes tend to
confuse between basic grammatical rules when speaking as well as writing (Hicks, 2013;
Trudgill & Hannah, 2013). Some common grammatical mistakes that have been identified in
American English include confused application of ‘its’ vs ‘it’s’, word level misunderstanding
between ‘fewer’ and ‘less’ or the misplaced use of ‘me’, ‘I’ and ‘myself’ (Hicks, 2013). One
mistake/misunderstanding that not many people focus on is the use of ‘who’ and ‘whom’ in a
sentence (Hinkel, 2013). A significant number of people have been recorded using ‘who’ in
place of ‘whom’ when referring to the object of a verb or a preposition (Lambek, 1997).
However, that should not be the case. This paper attempts to identify the usage of ‘whom’ in
two different scenarios – spoken (regular) and written (academic), in American English.
Based on the fact that in writing, more concentration can be given on the use of proper
grammar, the following hypothesis has been developed.
Hypothesis - Frequency of using ‘whom’ instead of ‘who’ as object of a verb or
preposition is higher in written than in spoken American English.
Sample and methodology
For the sample sentences, since I will be considering written and spoken instances of
American English, I will be considering the Corpus of Contemporary American English
(COCA). The database is huge and rich in American English language data for over twenty
years (Davies, 2010). The database has been divided neatly into various segments, however, I
will be looking at the ‘w_acad’ and the ‘w_spok’ segments which refer to the spoken and the
written aspects of American English respectively. The database has been divided year wise
and I will be considering 16 years of collected database (1991 – 2006) for the sample.
A sentence parsing algorithm based upon the Natural Language ToolKit (NLTK)
library is used to segregate each database into individual sentences (Rosovskaya, Roth &

2GRAMMAR
Srikumar, 2014). From there, the instances of ‘whom’ has been identified in each database
and analysed to identify the instances they are used in.
Results and discussion
In order to ease the process of analysis, I have divided the 32 databases (16 spoken
and 16 written) into 4 parts based on 4 consequent years each. The following section
discusses the results.
In the first part (1991 – 1994), we find that there are a total of 3,109 sentences in the
spoken segment and 2,074 sentences in the written segment. However, for the use cases of
‘whom’, a total of only 2 instances have been identified in the total number of sentences in
the spoken segment. In the written segment however, there are 5 instances of use of ‘whom’
in the total of 2074 sentences..
In the second part (1995-1998), for 3,472 sentences in the spoken segment, there are 3
cases of ‘whom’ used whereas for 2,288 sentences in the written segment, there are 4
instances.
In the third part (1999 - 2002), the spoken segment has 2833 sentences with only 3
instances of ‘whom’. The written segment, however, has 2086 sentences with 6 instances of
‘whom’ identified.
The final part (2003 – 2006) has a total of 3114 sentences in the spoken segment
where a total of 3 instances of ‘whom’ has been identified, whereas in the written segment,
for a total of 2182 sentences, there are 8 identified cases of ‘whom’.
While retrieving the data, since the instances of occurrence of ‘whom’ are less in
number, it was ensured that no instances where ‘whom’ is used to ask a question is recorded.
All instances of ‘whom’ identified is in the context of being an object of verb or preposition.
From the above identified data, the following inferences can be drawn:

3GRAMMAR
Firstly, the data sample retrieved from COCA records less instances of written
American English sentences than spoken. Secondly, less instances of ‘whom’ as object of
verb or preposition are recorded in the spoken segment than the written segment. In the
sixteen years of sample data that has been retrieved from the COCA database, there has been
a total of 11 instances where ‘whom’ was spoken in a total of 11,493 sentences. Alternately,
there has been a total of 23 instances in just 8630 sentences. This shows us that, the
mathematical probability of instances of ‘whom’ in spoken American English would be
approximately 953 instances per million sentences. In contrast, the written instances would be
approximately 2665 per million sentences. This shows that the occurrence of ‘whom’ as an
object of verb or preposition would be 64% higher in written American English than spoken
American English.
1991 - 1994 1995 - 1998 1999 - 2002 2003 - 2006
0
1
2
3
4
5
6
7
8
9
4-year distribution of spoken vs written 'whom'
whom' (spoken) whom' (written)

4GRAMMAR
953
2665
Occurrence of 'whom' in per million sentences estimation.
Spoken (per million) Written (per million)
Conclusion
The occurrence of ‘whom’ as recorded from the data proves that it is used more in
written American English than in Spoken American English. In written English, more
attention can be paid to the correct grammatical construction which cannot be done easily in
spoken context. Thus the hypothesis put forward by me is proven to be correct.

5GRAMMAR
References
Davies, M. (2010). The Corpus of Contemporary American English as the first reliable
monitor corpus of English. Literary and linguistic computing, 25(4), 447-464.
Hicks, W. (2013). English for journalists. Routledge.
Hinkel, E. (2013). Research findings on teaching grammar for academic writing. English
Teaching, 68(4), 3-21.
Lambek, J. (1997, September). Type grammar revisited. In International conference on
logical aspects of computational linguistics (pp. 1-27). Springer, Berlin, Heidelberg.
Rozovskaya, A., Roth, D., & Srikumar, V. (2014, April). Correcting grammatical verb errors.
In Proceedings of the 14th Conference of the European Chapter of the Association
for Computational Linguistics (pp. 358-367).
Trudgill, P., & Hannah, J. (2013). International English: A guide to the varieties of standard
English. Routledge.